DETAILED ACTION
This action is in response to the amendments filed on June 24th, 2026. A summary of this action:
Claims 1-5, 9-11, 13-15, 17-22, 24 have been presented for examination.
Claims 1-5, 9-11, 13-15, 17-22, 24 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea of both a mathematical concept and mental process without significantly more.
Claim(s) 1-2, 5, 9-11, 13, 15, 18-21, 24 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kabirzadeh, US 11,150,660 taken in view of Albrecht, Stefano V., et al. "Integrating planning and interpretable goal recognition for autonomous driving." arXiv preprint arXiv:2002.02277 (Feb. 6th, 2020).
Claim(s) 3-4 and 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kabirzadeh, US 11,150,660 taken in view of Albrecht, Stefano V., et al. "Integrating planning and interpretable goal recognition for autonomous driving." arXiv preprint arXiv:2002.02277 (Feb. 6th, 2020) in view of Kakade, Hrishikesh, et al. "Autonomous Highway Overtaking." Han University of Applied Sciences.Master’s (2018).
Claim(s) 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kabirzadeh, US 11,150,660 taken in view of Albrecht, Stefano V., et al. "Integrating planning and interpretable goal recognition for autonomous driving." arXiv preprint arXiv:2002.02277 (Feb. 6th, 2020) in view of Wu, Mo, and Benjamin Coifman. "Quantifying what goes unseen in instrumented and autonomous vehicle perception sensor data–A case study." Transportation research part C: emerging technologies 107 (2019): 105-119.
This action is Final
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Arguments/Amendments
Regarding the objections and 112 rejections
Withdrawn in view of amendment.
Regarding the § 101 Rejection
Maintained, updated as necessitated by amendment.
With respect to the prong 1 remarks, the first set of remarks, the Examiner notes that example 39 did not recite an abstract idea at all at prong 1 in the example’s analysis. The present claims recite an abstract idea, and recite additional elements that are not considered until prong 2 for they are not part of the abstract idea, rather additional to it.
With respect to the prong 2 remarks, they do not address the rationale of record at prong 2, nor do they identify any “technical challenge” but rather merely assert there is one. See MPEP § 2106.05(a) for the specification must provide the requisite details for an improvements consideration, the improvement must not just be a bare assertion of the improvement, and the claim must recite what provides the improvement expressly. And the improvement cannot be provided solely by the abstract idea, as MPEP § 2106.04(II)(A)(2): “Genetic Techs. Ltd. v. Merial LLC, 818 F.3d 1369, 1376, 118 USPQ2d 1541, 1546 (Fed. Cir. 2016) (eligibility "cannot be furnished by the unpatentable law of nature (or natural phenomenon or abstract idea) itself.")”
The Examiner does note page 19 of the specification for the “improved stack”, but this is expressly the result of a mental process: “The output of the test oracle 252 is informative to an expert 122 (team or individual), allowing them to identify issues in the stack 100 and modify the stack 100 to mitigate those issues (S 124 ).” – i.e. the person/people are doing the modification to the stack to improve it.
Examiner also notes page 53: “In this example, the goal recognition component 156 predicts multiple plausible trajectories 912 for a given vehicle and goal, rather than a single optimal trajectory This is beneficial because there are situations in which different trajectories may be (near)optimal but may lead to different predictions which could require different behaviour on the part of the ego vehicle” followed by math calculations with math equations in mathematical prose to provide this alleged improvement, i.e. solely provided by an abstract idea.
With respect to the 2B remarks, the remarks do not address the WURC evidence of record outside of a mere conclusory remark that the references “do not bridge this gap”. Furthermore, the 2B consideration for WURC does not consider the abstract idea itself, but rather what else besides for/beyond the abstract idea itself. MPEP § 2106.05(I): “ See also Alice Corp., 573 U.S. at 21-18, 110 USPQ2d at 1981 (citing Mayo, 566 U.S. at 78, 101 USPQ2d at 1968 (after determining that a claim is directed to a judicial exception, "we then ask, ‘[w]hat else is there in the claims before us?") (emphasis added)); RecogniCorp, LLC v. Nintendo Co., 855 F.3d 1322, 1327, 122 USPQ2d 1377 (Fed. Cir. 2017) ("Adding one abstract idea (math) to another abstract idea (encoding and decoding) does not render the claim non-abstract"). Instead, an "inventive concept" is furnished by an element or combination of elements that is recited in the claim in addition to (beyond) the judicial exception, and is sufficient to ensure that the claim as a whole amounts to significantly more than the judicial exception itself. Alice Corp., 573 U.S. at 27-18, 110 USPQ2d at 1981 (citing Mayo, 566 U.S. at 72-73, 101 USPQ2d at 1966)” – e.g. see Mayo in MPEP § 2106.05(d), e.g. see MPEP § 2106.05(a)(I) for BSG Tech LLC v. Buyseasons, Inc., 899 F.3d 1281, 1287-88, 127 USPQ2d 1688, 1693-94 (Fed. Cir. 2018).
Regarding the § 102/103 Rejection
Withdrawn, new grounds as necessitated by amendment.
With respect to the remarks regarding Kabirzadeh, see the rejection, in particular see that 1) the simulated vehicle has its own simulated controller, e.g. col. 21, ¶¶ 1-2: “The simulated 15 vehicle 416, however, can be configured to use a controller that is different than the controller used in vehicle 408”, wherein this controls the vehicle in the simulation, wherein the controllers of the vehicle and the object respectively provide their closed loop behaviors. Then, see col. 4, ¶¶ 3-4: “Thus, the simulated object can perform actions associated with the object as represented in the log data and/or perform actions that deviate from the object as represented in the log data. Thus, the simulated object can intelligently react to deviations in the behavior of the simulated vehicle” – e.g. in the pedestrian in the crosswalk example, it demonstrates that 1) the simulated object had closed-loop behavior reactive to the simulated vehicle, and 2) the simulated vehicle had closed-loop behavior reactive to the pedestrian, wherein col. 4, ¶¶ 3-4 clarify on the function of the controller to provide this result, and each of the vehicles in the simulation have their own controller. POSITA would have readily inferred that the behavior of the simulated vehicle, after the simulated object does an action, would be reactive to that action by the controller associated with the simulated vehicle.
The claim does not require that the non-ego agent actually have an action performed in the simulation in reaction to an action performed by the ego agent, but rather only that “behaviour” of the agents is determined. E.g. page 27: “For 5 example, with an ACC behaviour, target speeds may be set along the path which the agent will seek to match, but the agent decision logic 210 might be permitted to reduce the speed of the external agent below the target at any point along the path in order to maintain a target headway from a forward vehicle.” – a behaviour is first determined, and then an action is performed based on that behavior, but it may not be performed, e.g. the agent decision logic “might be permitted”, or it might not be permitted to do any action. Rather, it is just “autonomous decision-making functionality” of “some degree” (page 21, last paragraph), e.g. creating a planned route to later to be performed (page 22, ¶ 1).
The remarks imply that the claims expressly require that an action is performed by both the ego and non-ego vehicles reactive to each other, but that is not what is recited in the present claims.
To further clarify on this, the action is not required until what is now recited in dependent claim 5 (which previously only recited “with the aim of…” – i.e. it previously did not require this action to be performed. See new rejection of claim 5 below to clarify on how this is rejected in view of the prior art.
Furthermore, as was previously cited, see col. 21, ¶¶ 1-2, in particular: “At time T1 , the object 406 can represent a pedestrian crossing a crosswalk…The simulated 15 vehicle 416, however, can be configured to use a controller that is different than the controller used in vehicle 408. Therefore, at time T 2B in the simulated environment 412, the simulated vehicle 416 can come to a stop at a position (e.g., not within the intersection) that is farther away from the crosswalk and the simulated object 414. Additionally, time T 2B depicts the simulated object 418 coming to a stop such that it does not collide with simulated vehicle 416…In conducting the stop as depicted at time T 2B, the simulated vehicle 416 can perform a cost analysis. For example, the simulation component 314 can determine a cost associated with the simulated vehicle 416 applying a maximum braking (e.g., an emergency stop) which may result in excessive or uncomfortable accelerations for passengers represented in the simulated object 418” – i.e. the stopping action of the ego vehicle # 416 in 2B is reactive to the simulated object # 418 so as to avoid “excessive or uncomfortable accelerations for passengers
represented in the simulated object”, wherein the ego vehicle stops to also avoid hitting # 406 (the object in the crosswalk, e.g. a pedestrian), wherein the simulated object (non-ego vehicle) # 418 “coming to a stop such that it does not collide with simulated vehicle 416.” (reactive to the ego vehicle) as “However, the simulated model applied to the simulated object 418 allows the simulation component 314 to determine that the simulated vehicle 416 has come to a stop and prevent a collision by stopping at an earlier point than indicated by the log data 402.”
With respect to the remarks regarding Besse, see rejection below, as the amended subject matter has necessitated a new ground of rejection.
Claim Interpretation
With respect to the recitation of “closed loop”, this is given its BRI in view of page 4 last two paragraphs of the instant disclosure.
The term “stack” is interpreted in view of page 12, ¶¶ 2-3, including: “A stack can refer purely to software, i.e. one or more computer programs that can be executed on one or more general-purpose computer processors.”
Similar with the term ego agent, see page 12, last paragraph: “The ego agent is a real or simulated mobile robot that moves under the control of the stack under testing…”
Similar with the term trace, see page 14 ¶ 1: “A trace is a history of an agent's location and motion over the course of a scenario. There are many ways a trace can be represented…With regards to terminology, a "trace" and a "trajectory" may contain the same or similar types of information (such as a series of spatial and motion states over time). The term trajectory is generally favoured in the context of planning (and can refer to 10 future/predicted trajectories), whereas the term trace is generally favoured in relation to past behaviour in the context of testing/evaluation.”
Claim 18 recites the term substantially. This is interpreted in view of page 23 ¶ 2, and MPEP § 2173.05(b)(III)(D): “The term "substantially" is often used in conjunction with another term to describe a particular characteristic of the claimed invention. It is a broad term. In re Nehrenberg, 280 F.2d 161, 126 USPQ 383 (CCPA 1960). The court held that the limitation "to substantially increase the efficiency of the compound as a copper extractant" was definite in view of the general guidelines contained in the specification. In re Mattison, 509 F.2d 563, 184 USPQ 484 (CCPA 1975). The court held that the limitation "which produces substantially equal E and H plane illumination patterns" was definite because one of ordinary skill in the art would know what was meant by "substantially equal." Andrew Corp. v. Gabriel Electronics, 847 F.2d 819, 6 USPQ2d 2010 (Fed. Cir. 1988).”
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-5, 9-11, 13-15, 17-22, 24 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea of both a mathematical concept and mental process without significantly more.
As a point of clarity, the Examiner notes that the non-final action contained a minor typographical mistake wherein the extract a driving scenario limitation was not reproduced in the mental process rejection, however as was stated at page 21 of the non-final action: “The extracting is merely the final part of this mental process, e.g. writing down various pieces of information, e.g. make/model of the car, location of the car (e.g. referenced to a mile marker in visible range), etc.” – referring to this limitation, i.e. the limitation was addressed in the prior action expressly.
Step 1
Claim 19 is directed towards the statutory category of a process.
Claim 1 is directed towards the statutory category of an apparatus.
Claim 24 is directed towards the statutory category of an article of manufacture.
Claims 19 and 24, and the dependents thereof, are rejected under a similar rationale as representative claim 1, and the dependents thereof.
Step 2A – Prong 1
The claims recite an abstract idea of both a mental process and mathematical concept.
See MPEP § 2106.04: “...In other claims, multiple abstract ideas, which may fall in the same or different groupings, or multiple laws of nature may be recited. In these cases, examiners should not parse the claim. For example, in a claim that includes a series of steps that recite mental steps as well as a mathematical calculation, an examiner should identify the claim as reciting both a mental process and a mathematical concept for Step 2A Prong One to make the analysis clear on the record.”
To clarify, see the USPTO 101 training examples, available at https://www.uspto.gov/patents/laws/examination-policy/subject-matter-eligibility.
The mathematical concept recited in claim 1 is:
for each of the goals or behaviours, determining an expected trajectory model, wherein the expected trajectory model simulates future behaviour based on the goal or behaviour;
…probabilistically by… for each of the goals or behaviours, determining an expected trajectory model, wherein the expected trajectory model simulates future behaviour based on the goal or behaviour;… comparing the observed trace of the real-world agent with the expected trajectory model for each of the goals or behaviours, to determine a likelihood of each of the goals or behaviours, thus determining a distribution over the goals or behaviours – math calculations/equations/relationships in textual form. Page 42, ¶¶ 4-5: “The expected trajectory model may simply be a (single) predicted for a given goal [a mental judgement/evaluation], but in the present examples it takes the form of a predicted trajectory distribution [a math concept that is simple enough for a person to mentally to do it as well] for the goal in question.” – see fig. 8A and 8B to clarify; and subsection F starting on page 50 to clarify on the trajectory distribution being a math concept expressed as mathematical equations, in particular eq. 6 and its later definition of “L”, followed by another mental process and math concept (the “determine a likelihood” is a math calculation in textual form)
To clarify on the comparing with this – page 44 ¶ 2: “In other words, the goal recognition component 156 predicts, for each of the hypothesized 5 goals, a set of one or more possible trajectories that the other vehicle might have taken in the time interval 11t and a likelihood of each trajectory, on the assumption that the other vehicle was executing that goal during that time period (i.e. what the other vehicle might have done during time interval 11t had it been executing that goal).This is then compared with the actual trace of the other vehicle within that time period (i.e. what the other vehicle actually 10 did), to determine a likelihood of each goal for the time period 11” – in view of pages 41-42 as cited above), i.e. this is a mental step, i.e. two goals (page 41: "follow lane" and "switch lane"), each goal with one possible trajectory, so observe the real trajectory (e.g. the vehicle is moving into the other lane of traffic) and thus determine the likelihood of it going to that goal is 1 (also, provide more information, e.g. more observed trajectories in this case, and it would likely merely go down to a coin flip, i.e. a 50/50 likelihood of one of the two goals).
Under the broadest reasonable interpretation, the claim recites a mathematical concept – the above limitations are steps in a mathematical concept such as mathematical relationships, mathematical formulas or equations, and mathematical calculations. If a claim, under its broadest reasonable interpretation, is directed towards a mathematical concept, then it falls within the Mathematical Concepts grouping of abstract ideas. In addition, as per MPEP § 2106.04(a)(2): “It is important to note that a mathematical concept need not be expressed in mathematical symbols, because "[w]ords used in a claim operating on data to solve a problem can serve the same purpose as a formula." In re Grams, 888 F.2d 835, 837 and n.1, 12 USPQ2d 1824, 1826 and n.1 (Fed. Cir. 1989). See, e.g., SAP America, Inc. v. InvestPic, LLC, 898 F.3d 1161, 1163, 127 USPQ2d 1597, 1599 (Fed. Cir. 2018)”
See MPEP § 2106.04(a)(2).
To clarify, see the USPTO 101 training examples, available at https://www.uspto.gov/patents/laws/examination-policy/subject-matter-eligibility.
The mental process recited in claim 1 is:
process real-world driving data to extract therefrom at least one observed trace of a real-world agent within a road layout, the observed trace having spatial and motion components; - see page 20: “Data 140 of a real-world run is passed to a 'ground-trothing' pipeline 142 (trace extraction component) for the purpose of generating scenario ground truth… The run data is processed within the ground truthing pipeline 142, in order to generate appropriate ground truth 144 (trace(s) and contextual data) for the real-world run. The ground truth of the real-world run 144 comprises an extracted ego trace of the ego agent and one or more extracted agent trace(s) of one or more other (non-ego) agents. As discussed, the ground-truthing process could be based on manual annotation of the 'raw' run data 142, or the process could be entirely automated (e.g. using offline perception method(s)), or a combination of manual and automated ground truthing could be used.”
See page 13 to clarify: “In a real-world scenario run, a "perfect" representation of the
scenario run does not exist in the same sense; nevertheless, suitably informative ground truth
can be obtained in numerous ways, e.g. based on manual annotation of on-board sensor data,”
Page 27, second to last paragraph: “For example, each trace 212a, 212b may take the form of a spatial path having motion data associated with points along the path such as speed, acceleration, jerk (rate of change of acceleration), snap (rate of change of jerk) etc.”
E.g. the mental process of this step would be a person simply observing measured data, e.g. video from a dash cam (or similar such camera), and mentally evaluating it to determine an approximate trajectory of another vehicle, e.g. observing police dash cam footage on the nightly news, and evaluating/judging that the other vehicle is about to go in a certain direction.
apply at least one of goal recognition and behaviour recognition to the at least one observed trace, to infer a goal or behaviour of the real-world agent within the road layout, and extract a driving scenario defining a version of the road layout and at least one non-ego agent to be simulated, the non-ego agent associated with the inferred goal or behaviour for implementing in simulation within the defined road layout … by:;
A mental process. See page 14 ¶ 1: “A trace is a history of an agent's location and motion over the course of a scenario.” See pages 22-23: “The present techniques can also be applied with deterministic goal recognition, in which case the goal sampling component 151 may be omitted, and deterministic goals may be inferred and provided directly to the agent decision logic 210… A goal generally refers to a relatively longer-term objective (e.g. which might remain fixed over the course of a simulated run), whilst a maneuver generally occurs over a relatively shorter time scale.”. Page 40, second to last paragraph: “A goal may for example be captured as a desired location (reference point) on a map, which
the ego vehicle is attempting to reach from a current location on the map. For example the
desired location may be defined in relation to a particular junction, lane layout, roundabout
exit etc. The map, in this context, refers to the static road layout of a scenario description
201a…” and page 41, last paragraph: “For example, given a set 25 of non-ego vehicles in the vicinity of a road junction, roundabout or other road layout indicated on the map (the driving context), suitable goals may be hypothesised from the road layout alone (without taking into account any observed historical behaviour of the agent). By way of example, if the other vehicle is currently driving on a multi-lane road, with no nearby junctions, the set of hypothesised goals may consist of "follow lane" and "switch lane". As 30 another example, with a set of non-ego agents in the vicinity of a left-tum junction, the hypothesised goals may be tum left and continue straight. As indicated, such goals are defined with reference to suitable reference points on the map.”
Page 51: “A heuristic function is used to generate a set of possible goals qi for vehicle i based on its location and context information such as road layout and traffic rules. The goal recognition component 156 defines one goal for the end of the vehicle's current road and goals for end of each reachable connecting road, bounded by the ego vehicle's view region. Infeasible goals, such as locations behind the vehicle, are not included.”
To clarify, given the generality recited herein, this is a mental process of a person mentally observing the trajectory of a vehicle, and simply observing a behavior, e.g. that the vehicle is driving aggressively/recklessly fast, or mentally judging when the vehicle is to go, e.g. observing that although the driver in front of them does not have their blinker on, the driver’s trajectory is indicating (e.g. by speeding up and starting to drift in a different lane) that they are going to do a lane change (a goal), or similarly observing a vehicle, e.g. an 18 wheeler, is moving on a trajectory to an off-ramp on I-70 in the middle of rural Kansas, with the other rest stop being a trucking friendly gas station, so the person infers that the goal of the 18-wheeler is to stop for gas at that gas station. Or similarly they observe an ambulance with its lights on speeding in a trajectory that would go in the general direction of the hospital, thereby they are readily able to infer its destination/goal is the hospital.
Or observing a mini-van on Saturday morning, driving on a trajectory down a road near the local athletic fields, and upon further observing the soccer team stickers on the back of the mini-van, inferring that the goal is likely the nearby soccer fields.
The extracting is merely the final part of this mental process, e.g. writing down various pieces of information, e.g. make/model of the car, location of the car (e.g. referenced to a mile marker in visible range), etc.
determining a set of goals or behaviours for the real-world agent; - a mental process such as a mental evaluation/judgement/opinion. Page 41, last paragraph: “…suitable goals may be hypothesised from the road layout alone (without taking into account any observed historical behaviour of the agent)….By way of example, if the other vehicle is currently driving on a multi-lane road, with no nearby junctions, the set of hypothesised goals may consist of "follow lane" and "switch lane".”)
for each of the goals or behaviours, determining an expected trajectory model, wherein the expected trajectory model simulates future behaviour based on the goal or behaviour; and comparing the observed trace of the real-world agent with the expected trajectory model for each of the goals or behaviours, to determine a likelihood of each of the goals or behaviours, thus determining a distribution over the goals or behaviours, and extract a driving scenario defining a version of the road layout and at least one non-ego agent to be simulated, the non-ego agent associated with an inferred goal or behaviour for implementing in simulation within the defined version of the road layout; - a mental process, wherein the math, as discussed above, is simple enough for a person to mentally evaluate such as with physical aids. Clarification in above discussion of these limitations.
extract a driving scenario defining a version of the road layout and at least one non-ego agent to be simulated, the non-ego agent associated with an inferred goal or behaviour for implementing in simulation within the defined version of the road layout - The extracting is merely the final part of this mental process that was discussed above, e.g. writing down various pieces of information, e.g. make/model of the car, location of the car (e.g. referenced to a mile marker in visible range), etc.
Under the broadest reasonable interpretation, these limitations are process steps that cover mental processes including an observation, evaluation, judgment or opinion that could be performed in the human mind or with the aid of physical aids but for the recitation of a generic computer component. If a claim, under its broadest reasonable interpretation, covers a mental process but for the recitation of generic computer components, then it falls within the "Mental Process" grouping of abstract ideas. A person would readily be able to perform this process either mentally or with the assistance of physical aids. See MPEP § 2106.04(a)(2).
To clarify, see the USPTO 101 training examples, available at https://www.uspto.gov/patents/laws/examination-policy/subject-matter-eligibility. In particular, with respect to the physical aids, see example # 45, analysis of claim 1 under step 2A prong 1, including: “Note that even if most humans would use a physical aid (e.g., pen and paper, a slide rule, or a calculator) to help them complete the recited calculation, the use of such physical aid does not negate the mental nature of this limitation.”; also see example # 49, analysis of claim 1, under step 2A prong 1: “Moreover, the recited mathematical calculation is simple enough that it can be practically performed in the human mind. Even if most humans would use a physical aid, like a pen and paper or a calculator, to make such calculations, the use of a physical aid would not negate the mental nature of this limitation.”
As such, the claims recite an abstract idea of both a mental process and mathematical concept.
Step 2A, prong 2
The claimed invention does not recite any additional elements that integrate the judicial exception into a practical application. Refer to MPEP §2106.04(d).
The following limitations are merely reciting the words "apply it" (or an equivalent) with the judicial exception, or merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea, as discussed in MPEP § 2106.05(f), including the “Use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application or provide significantly more”:
Preambles of the independent claims recite mere instructions to do it on a computer with generic computer components
The following limitations are considered as generally linking to a particular technological environment of simulating the scenario, as mere instructions to “apply it” with results-oriented functional language that provides no details on how this simulation is to be performed in a technological manner nor how the agent decision logic is to actually do the implementing in a technological manner (do note: page 27, ¶ 1: “target speeds may be set along the path which the agent will seek to match, but the agent decision logic 210 might be permitted to reduce the speed of the external agent below the target at any point along the path in order to maintain a target headway from a forward vehicle.” - people routinely do this mentally and are legally required to when driving, i.e. maintaining a safe braking distance from the car in front of them); and furthermore this is an insignificant application/insignificant computer implementation of the abstract idea:
run a simulation based on the extracted driving scenario, in which an ego agent and the non-ego agent each exhibit closed-loop behaviour, wherein the closed-loop behaviour of the ego agent is determined by autonomous decisions taken in the AV stack under testing in response to simulated inputs, reactive to the non-ego agent;
and determine the closed-loop behaviour of the non-ego agent in the simulation, by applying agent decision logic to implement the inferred goal or behaviour, reactive to the ego agent.
A claim that integrates a judicial exception into a practical application will apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception, such that the claim is more than a drafting effort designed to monopolize the judicial exception. See MPEP § 2106.04(d).
MPEP 2106.04(II)(A)(2) “…Instead, under Prong Two, a claim that recites a judicial exception is not directed to that judicial exception, if the claim as a whole integrates the recited judicial exception into a practical application of that exception. Prong Two thus distinguishes claims that are "directed to" the recited judicial exception from claims that are not "directed to" the recited judicial exception…Because a judicial exception is not eligible subject matter, Bilski, 561 U.S. at 601, 95 USPQ2d at 1005-06 (quoting Chakrabarty, 447 U.S. at 309, 206 USPQ at 197 (1980)), if there are no additional claim elements besides the judicial exception, or if the additional claim elements merely recite another judicial exception, that is insufficient to integrate the judicial exception into a practical application. See, e.g., RecogniCorp, LLC v. Nintendo Co., 855 F.3d 1322, 1327, 122 USPQ2d 1377 (Fed. Cir. 2017) ("Adding one abstract idea (math) to another abstract idea (encoding and decoding) does not render the claim non-abstract"); Genetic Techs. Ltd. v. Merial LLC, 818 F.3d 1369, 1376, 118 USPQ2d 1541, 1546 (Fed. Cir. 2016) (eligibility "cannot be furnished by the unpatentable law of nature (or natural phenomenon or abstract idea) itself."). For a claim reciting a judicial exception to be eligible, the additional elements (if any) in the claim must "transform the nature of the claim" into a patent-eligible application of the judicial exception, Alice Corp., 573 U.S. at 217, 110 USPQ2d at 1981, either at Prong Two or in Step 2B” and MPEP § 2106(I): “Mayo, 566 U.S. at 80, 84, 101 USPQ2dat 1969, 1971 (noting that the Court in Diamond v. Diehr found “the overall process patent eligible because of the way the additional steps of the process integrated the equation into the process as a whole,”” – and see MPEP § 2106.05(e).
To further clarify, MPEP § 2106.04(II)(A)(1): “Alice Corp., 573 U.S. at 216, 110 USPQ2d at 1980 (citing Mayo, 566 US at 71, 101 USPQ2d at 1965). Yet, the Court has explained that ‘‘[a]t some level, all inventions embody, use, reflect, rest upon, or apply laws of nature, natural phenomena, or abstract ideas,’’ and has cautioned ‘‘to tread carefully in construing this exclusionary principle lest it swallow all of patent law” See also Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 1335, 118 USPQ2d 1684, 1688 (Fed. Cir. 2016) ("The ‘directed to’ inquiry, therefore, cannot simply ask whether the claims involve a patent-ineligible concept, because essentially every routinely patent-eligible claim involving physical products and actions involves a law of nature and/or natural phenomenon").”
As a point of clarity, RecogniCorp, LLC v. Nintendo Co., 855 F.3d 1322, 1327, 122 USPQ2d 1377 (Fed. Cir. 2017) ("Adding one abstract idea (math) to another abstract idea (encoding and decoding) does not render the claim non-abstract"); Genetic Techs. Ltd. v. Merial LLC, 818 F.3d 1369, 1376, 118 USPQ2d 1541, 1546 (Fed. Cir. 2016) (eligibility "cannot be furnished by the unpatentable law of nature (or natural phenomenon or abstract idea) itself." discussed in MPEP § 2106.04(II)(A)(2) as well as MPEP § 2106.04(I): “Synopsys, Inc. v. Mentor Graphics Corp., 839 F.3d 1138, 1151, 120 USPQ2d 1473, 1483 (Fed. Cir. 2016) ("a new abstract idea is still an abstract idea") (emphasis in original).
The claimed invention does not recite any additional elements that integrate the judicial exception into a practical application. Refer to MPEP §2106.04(d).
Step 2B
The claimed invention does not recite any additional elements/limitations that amount to significantly more.
The following limitations are merely reciting the words "apply it" (or an equivalent) with the judicial exception, or merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea, as discussed in MPEP § 2106.05(f), including the “Use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application or provide significantly more”:
Preambles of the independent claims recite mere instructions to do it on a computer with generic computer components, and the simulator of claim 19.
The following limitations are considered as generally linking to a particular technological environment of simulating the scenario, as mere instructions to “apply it” with results-oriented functional language that provides no details on how this simulation is to be performed in a technological manner nor how the agent decision logic is to actually do the implementing in a technological manner (do note: page 27, ¶ 1: “target speeds may be set along the path which the agent will seek to match, but the agent decision logic 210 might be permitted to reduce the speed of the external agent below the target at any point along the path in order to maintain a target headway from a forward vehicle.” - people routinely do this mentally and are legally required to when driving, i.e. maintaining a safe braking distance from the car in front of them); and furthermore this is an insignificant application/insignificant computer implementation of the abstract idea:
run a simulation based on the extracted driving scenario, in which an ego agent and the non-ego agent each exhibit closed-loop behaviour, wherein the closed-loop behaviour of the ego agent is determined by autonomous decisions taken in the AV stack under testing in response to simulated inputs, reactive to the non-ego agent;
and determine the closed-loop behaviour of the non-ego agent in the simulation, by applying agent decision logic to implement the inferred goal or behaviour, reactive to the ego agent.
In addition, the above insignificant extra-solution activities are also considered as well-understood, routine, and conventional activities, as discussed in MPEP § 2106.05(d):
Curiel-Ramirez, Luis A., et al. "Hardware in the loop framework proposal for a semi-autonomous car architecture in a closed route environment." International Journal on Interactive Design and Manufacturing (IJIDeM) 13.4 (2019): 1647-1658. § 2, incl. § 2.1: “The creation of simulation tools have become essential for the development of this type of technologies for autonomous vehicles (AV). These simulators have been created with different approaches and objectives, such as for the training of machine learning systems applied to control or vision systems; others for the mapping, location, path planning and sensor fusion of the vehicle. In this section we will present the state of the art of some of them, which were used for the development of certain implementations of the work. The simulation tools of autonomous vehicles have become very important in recent years. These simulators allowto simplify and optimize the systems of vision, control, mapping, location and other blocks that compose the autonomous vehicles. In the same way there are simulators of a more general use that allow to visualize and manage part of the systems that make up the autonomous vehicles” – then, see the other subsections of section 2 which discuss these various conventional simulators, e.g. § 2.1.1: “CARLA™ [9] is an open-source simulator for autonomous driving research. CARLA™ have been developed with the objective of working as a development, training, and validation of autonomous urban driving systems (Fig. 3). In addition to open-source code and protocols, CARLA™ provides open digital assets (urban layouts, buildings, vehicles) that were created for this purpose and can be used freely. The simulation platform supports flexible specification of sensor suites and environmental conditions. CARLA™ have been used to study the performance of three approaches to autonomous driving: a classic modular pipeline, an end-to end model trained via imitation learning, and an end-to-end model trained via reinforcement learning”, e.g. § 2.1.2: “…Developers can create accurate, detailed models of both systems and environments, providing them with intelligence Fig. 3 CARLA™ autonomous vehicles simulator [17] by using methods such as deep learning, imitation learning and reinforcement learning. Tools such as Bonsai [6] can be used to train the models across a variety of environmental conditions and vehicle scenarios in the cloud, on Microsoft Azure, much faster and safer than is feasible in the real world. After training is complete, designers can deploy these trained models onto actual hardware [34].” – etc. see the other subsections for further clarification. See § 2.2 for more clarification as well.
Kakade, Hrishikesh, et al. "Autonomous Highway Overtaking." Han University of Applied Sciences. Master’s (2018). § 2.5, followed by § 5.1.
Codevilla, Felipe, et al. "Exploring the limitations of behavior cloning for autonomous driving." Proceedings of the IEEE/CVF international conference on computer vision. 2019. Abstract and § 1, including the third to last paragraph, then see § 2.
Feng, Shuo, et al. "Testing scenario library generation for connected and automated vehicles, part I: Methodology." IEEE Transactions on Intelligent Transportation Systems 22.3 (2020): 1573-1582. § I ¶¶ 1-4
Feng, Shuo, et al. "Testing scenario library generation for connected and automated vehicles, part II: Case studies." IEEE Transactions on Intelligent Transportation Systems 22.9 (2020): 5635-5647.§ I, then see § III.C, then see page 11, col. 1, ¶ 2.
Fremont, Daniel J., et al. "Formal scenario-based testing of autonomous vehicles: From simulation to the real world." 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC). IEEE, 2020. § I incl. subsection related work. § II.C.
Huang, Xin, et al. "Online risk-bounded motion planning for autonomous vehicles in dynamic environments." Proceedings of the International Conference on Automated Planning and Scheduling. Vol. 29. 2019. Sections on “Approach” and “Intention-Aware Risk-Bounded Motion Planning” and algorithms 1-2. Then see the “Experiments” section on page 219, and fig. 3 and 5.
The claimed invention is directed towards an abstract idea of both a mathematical concept and a mental process without significantly more.
Regarding the dependent claims
Claim 2 is adding more mere instructions to do an abstract idea on a computer (the test oracle), with a mere data gathering step (the receiving step), and a mental process given the generality of the other steps recited, i.e. observe a trajectory/trace, and mentally compared (by judgement/evaluation) it against performance metrics to score it, e.g. observe the “acceleration” merely output from the simulation (page 31, last paragraph), and compare it to some mental threshold for the max comfortable acceleration, e.g. a numerical threshold value such as 3.5 m/s^2. To clarify, page 31, second to last paragraph: “The performance metrics 254 can be based on various factors, such as distance speed etc. In the described system, these can mirror a set of applicable road rules, such as the Highway Code applicable to road users in the United Kingdom.” – see MPEP § 2106.04(a)(2)(III)(C): “Another example is FairWarning IP, LLC v. Iatric Sys., Inc., 839 F.3d 1089, 120 USPQ2d 1293 (Fed. Cir. 2016). The patentee in FairWarning claimed a system and method of detecting fraud and/or misuse in a computer environment, in which information regarding accesses of a patient’s personal health information was analyzed according to one of several rules (i.e., related to accesses in excess of a specific volume, accesses during a pre-determined time interval, or accesses by a specific user) to determine if the activity indicates improper access. 839 F.3d. at 1092, 120 USPQ2d at 1294. The court determined that these claims were directed to a mental process of detecting misuse, and that the claimed rules here were "the same questions (though perhaps phrased with different words) that humans in analogous situations detecting fraud have asked for decades, if not centuries." 839 F.3d. at 1094-95, 120 USPQ2d at 1296.”
Claim 3 – mere data outputting that is WURC in view of example 46, claim 1 for its displaying step WURC analysis; also see additional evidence in MPEP § 2106.05(d)(II)
Claim 4 – merely further limiting the mental process, and adding another step in it (e.g. observe a chart of data, and observe where values of the chart exceed a threshold value)
Claim 5 – rejected under a similar rationale as the applying limitation as discussed above, i.e. this is further merely expressing a desired result in purely functional language with no recitation of how to achieve this result
Claim 9 – further limiting the abstract idea of claim 1 above
Claim 10 – rejected under a similar rationale as claim 1 above, including being a mental step for the prediction of the first trajectory model given the generality recited, as well as a simple mental comparison
Claim 11 – adding a math calculation in textual form, and mentally comparing the results of the calculation. See page 49, last two paragraphs to clarify.
Claim 13 – math calculations in textual form, for similar reasons as discussed above in view of MPEP § 2106.04(a)(2)(III)(C) for SAP v. InvestPic, followed by merely adding this information to later be used in the token post-solution activity (see above discussion in independents for the agent decision logic)
Claim 14 – mere data outputting with generic computer technology (the GUI) recited in a high level of generality – WURC in view of example 46, claim 1, step 2B for its displaying step and MPEP § 2106.05(d)(II)
Claim 15 merely further limiting the mental process as discussed above
Claim 17 – mere instructions to do it on a computer, in view of MPEP § 2106.05(f): “Other examples where the courts have found the additional elements to be mere instructions to apply an exception, because they do no more than merely invoke computers or machinery as a tool to perform an existing process include: i. A commonplace business method or mathematical algorithm being applied on a general purpose computer, Alice Corp. Pty. Ltd. V. CLS Bank Int’l, 573 U.S. 208, 223, 110 USPQ2d 1976, 1983 (2014); Gottschalk v. Benson, 409 U.S. 63, 64, 175 USPQ 673, 674 (1972); Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015);”
Claim 18 – purely results oriented, so see the rationale above for the agent decision logic
Claims 20 – rejected under a similar rationale as its parallel claims above
Claim 21 – a mental process. Page 19, ¶ 2: “The output of the test oracle 252 is informative to an expert 122 (team or individual), allowing them to identify issues in the stack 100 and modify the stack 100 to mitigate those issues (S 124). The results also assist the expert 122 in selecting further 20 scenarios for testing (S 126), and the process continues, repeatedly modifying, testing and evaluating the performance of the stack 100 in simulation”
Claim 22 – generally linking to the technological environment of machine learning, and an insignificant application WUC in view of the above evidence. To clarify, the repeat evaluations is a mental process in view of Page 19, ¶ 2: “The output of the test oracle 252 is informative to an expert 122 (team or individual), allowing them to identify issues in the stack 100 and modify the stack 100 to mitigate those issues (S 124). The results also assist the expert 122 in selecting further 20 scenarios for testing (S 126), and the process continues, repeatedly modifying, testing and evaluating the performance of the stack 100 in simulation. The improved stack 100 is eventually incorporated (S 125) in a real-world AV 101, equipped with a sensor system 110 and an actor system 112.”
The claimed invention is directed towards an abstract idea of both a mathematical concept and a mental process without significantly more.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1-2, 5, 9-11, 13, 15, 18, 19-21, 24 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kabirzadeh, US 11,150,660 taken in view of Albrecht, Stefano V., et al. "Integrating planning and interpretable goal recognition for autonomous driving." arXiv preprint arXiv:2002.02277 (Feb. 6th, 2020).
Albrecht qualifies as prior art under § 102(a)(1) as per MPEP § 2153.01(a): “ If, however, the application names fewer joint inventors than a publication (e.g., the application names as joint inventors A and B, and the publication names as authors A, B and C), it would not be readily apparent from the publication that it is an inventor-originated disclosure and the publication would be treated as prior art under AIA 35 U.S.C. 102(a)(1) unless there is evidence of record that an exception under AIA 35 U.S.C. 102(b)(1) applies”
Examiner suggests, to address this rejection, a similar evidentiary showing as discussed MPEP § 2155.01: “Example 1… In response to the prior art rejection, AB Corp.'s attorney files a declaration under 37 CFR 1.130(a) signed by Anwesha averring that she is the sole inventor of the subject matter X disclosed in the article. Anwesha also explains in the declaration that Bob was a graduate student working under her direction and supervision and did not contribute to the conception of the claimed invention. In other words, Anwesha’s declaration establishes that Bob was not a joint inventor of subject matter X. The examiner should withdraw the rejection because the declaration establishes that the journal article is not prior art. Anwesha's declaration includes a statement that Anwesha invented X, so there is no need for the examiner to require a signed inventor's oath or declaration under 37 CFR 1.63 at this time in order to withdraw the rejection. Furthermore, the declaration properly includes a reasonable explanation of Bob's involvement with the journal article. A statement from Bob is not needed. There is no evidence in the record to suggest that Bob was a joint inventor of X.” – see remaining examples to further clarify.
Regarding Claim 1
Kabirzadeh teaches:
A computer system for testing an autonomous vehicle (AV) stack in simulation, the computer system comprising: memory configured to store computer-readable instructions; and one or more hardware processors coupled to the memory and configured to execute the computer-readable instructions, which upon execution cause the computer system to: (Kabirzadeh, abstract and cf. 1-3 along with accompanying description)
process real-world driving data to extract therefrom at least one observed trace of a real-world agent within a road layout, the observed trace having spatial and motion components; (Kabirzadeh, as cited above, e.g. cf. 3, # 304, # 310, and # 308, see fig. 5 # 502 and # 510 to clarify, then cf. 7 # 702-706, see col. 7-8 to clarify incl.: “The vehicle(s) 104 can include a computing device that includes a perception engine and/or a planner and perform operations such as detecting, identifying, segmenting, classifying, and/or tracking objects from sensor data collected from the environment 102… The vehicle computing device can use the sensor data to
generate a trajectory [example of a trace] for the vehicle(s) 104…” – see the paragraph split between the columns in particular, and col. 8 ¶ 2
apply at least one of goal recognition and behaviour recognition to the at least one observed trace, to infer a goal or behaviour of the real-world agent within the road layout, and extract a driving scenario defining a version of the road layout and at least one non-ego agent to be simulated, the non-ego agent associated with an inferred goal or behaviour for implementing in simulation within the defined version of the road layout…; (Kabirzadeh, cf. 3, # 324-326 – to clarify, col. 13, ln. 55-65: “For the purpose of this discussion, a route can be a sequence of waypoints for traveling between two locations. As non-limiting examples, waypoints include streets, intersections, global positioning system (GPS) coordinates, etc.”, and col. 4-5 paragraph split between: “…In some examples, the SES can determine waypoint(s) [example of a goal inferred] and/or paths associated with a simulated object. For purposes of illustration only such waypoints or paths can be based at least in part on log data representing the object corresponding to the simulated object. In some examples, waypoints can be determined based on a curvature of a path segment or can be added manually in a user interface, discussed herein. Such waypoints or paths can be associated with various costs or weights that influence a behavior [example of behavior inferred] of the simulated object in the simulated environment… Such waypoints or paths can be associated with events or actions, such as a lane change action. For example, the log data can indicate that an object performed a lane change action that placed the object to the right of the vehicle. Additionally, the log data can indicate that the vehicle attempted to perform a lane change action to the right but did not complete the action due to the object. In order to preserve this interaction in the simulated scenario, the SES can assign a waypoint associated with the trajectory of the simulated object such that the simulated object performs the lane change action and remains on the right side of the simulated vehicle.”
To further clarify, see col. 15-16, incl.: “The waypoint attributes component 242 can determine attributes associated with a waypoint… The waypoint attributes component 242 can determine attributes such as, for example, a speed, a steering angle, or a time associated with the waypoint (e.g., to enforce the simulated object to simulate a speed and/or a steering angle at a waypoint in the simulated environment, to arrive at a waypoint at a specified time in the simulated scenario ( or within a period of time), and/or to perform a specified action at a specified time in the simulated scenario ( or within a period of time)). In another example, a waypoint may enforce a location of a simulated object at a particular time in the simulated data. Of course, such waypoints may be associated with any one or more additional parameters, such as, for example, yaw rates, accelerations, predicted trajectories, uncertainties, and any other data associated with entities in the data. In some examples, a waypoint can be associated with various weight(s) that represent an "importance" of enforcing a behavior of the simulated object with respect to the waypoint. In some examples, the weight(s) associated with a waypoint can influence a cost associated with the simulated object deviating from or adhering to the waypoint behavior… For purposes of illustration only, the log data can indicate that the object traversed a trajectory that included executing a turn at 5 mph and with a steering angle of30 degrees. The waypoint attributes component 242 can determine a waypoint associated with the simulated object that has attributes of a speed of 5 mph and a steering angle of 30 degrees at the location as represented in the log data. As the computing device(s) 232 executes the simulated scenario, the waypoints can guide the simulated object in the simulated environment by applying the attributes associated with the waypoints to the simulated object…”
Col. 20, ¶ 4 starting at ln. 31 to further clarify: “The waypoints attributes component 242 can determine the first waypoint 324 at location G) and the second waypoint 326 at location@. When the computing device(s) 312 executes the simulated scenario, the computing device(s) 35 312 can generate a trajectory that respects or otherwise represents the attributes associated with the waypoints 324 and/or 326 (e.g., by minimizing a difference between a trajectory of the object in simulation and the observed trajectory of the object). For example, the simulated object 40 318 can traverse the simulated environment 316 such that, when the simulated object 318 reaches the first waypoint 324, the behavior of the simulated object 318 is substantially similar (at least in part and within a threshold) to the attributes associated with the first waypoint 324…”
To clarify, the Examiner notes the ego vehicle of Kabirzadeh is the “Simulated vehicle”, e.g. # 416, corresponding to the “vehicle”, e.g. # 408, and the non-ego vehicle is called the “object” (# 410)/”Simulated object” (# 418)
To clarify on behavior, also see col. 17, last paragraph: “In some instances, the scenario component 246 can determine, based on behavior data in the log data, that an object as an aggressive object, a passive object, a neutral object, and/or other types of behaviors and apply behavior instructions associated with the behavior (e.g., a 65 passive behavior, a cautious behavior, a neutral behavior, and/or an aggressive behavior) to the simulated object”
…run a simulation based on the extracted driving scenario, in which an ego agent and the non-ego agent each exhibit closed-loop behaviour, wherein the closed-loop behaviour of the ego agent is determined by autonomous decisions taken in the AV stack under testing in response to simulated inputs, reactive to the non-ego agent; and determine the closed-loop behaviour of the non-ego agent in the simulation, by applying agent decision logic to implement the inferred goal or behaviour, reactive to the ego agent. (Kabirzadeh, as was cited above, then see:
col. 4, ¶ 4: “For purposes of illustration only, in a scenario where a simulated object is following a simulated vehicle, if the simulated vehicle begins braking sooner or more aggressively than the braking behavior represented in the log data, the simulated object may modify its trajectory by braking sooner or changing lanes, for example, to avoid a collision or near-collision with the simulated vehicle” and col. 17, ¶ 3: “…In some instances, the scenario
component 246 can identify a simulated pedestrian model and apply it to the simulated objects associated with pedestrians. In some instances, the simulated object models can 45 use controllers that allow the simulated objects to react to the simulated environment and other simulated objects (e.g., modeling physics-based behaviors and incorporating collision checking). For example, a simulated vehicle object can stop at a crosswalk if a simulated pedestrian crosses the 50 crosswalk as to prevent the simulated vehicle from colliding with the simulated pedestrian.”, e.g. col. 21 ¶¶ 1-2: “…The simulated 15 vehicle 416, however, can be configured to use a controller that is different than the controller used in vehicle 408. Therefore, at time T 2B in the simulated environment 412, the simulated vehicle 416 can come to a stop at a position (e.g., not within the intersection) that is farther away from the crosswalk and the simulated object 414. Additionally, time T 2B depicts the simulated object 418 coming to a stop such that it does not collide with simulated vehicle 416. By way of example, if the motion of simulated object 418 relied solely on the log data 402, the simulated object 418 would have collided with the simulated vehicle 416. However, the simulated model applied to the simulated object 418 allows the simulation component 314 to determine that the simulated vehicle 416 has come to a stop [i.e. reactive to the ego vehicle] and prevent a collision by stopping at an earlier point than indicated by the log data 402… Additionally, the simulation component 314 can determine a cost associated with the simulated vehicle 416 applying a minimum braking force, which may result in a collision or near-collision with the simulated object 418. The simulation component 314 can, using cost minimization algorithms (e.g., gradient descent), determine a trajectory to perform the stop in a safe manner while minimizing or optimizing costs.” – to clarify on this implementing the waypoints, see waypoints # 1-3 in fig. 5, and col. 22 ¶¶ 2-3: “The waypoint attributes component 242 can identify three
different waypoints e.g., (1, 2, 3) and the scenario component 246 can generate three different simulated environments 514, 516, and 518… The simulated object 520(1)-(3) traverses trajectory 522 (1)-(3) to avoid simulated vehicle 524(1)-(3). Each trajectory 522(1)-(3) can be associated with a cost and the controller associated with the simulated object 520(1)-(3) 25 can determine the trajectory to minimize the cost”
To clarify, see the abstract: “The scenarios can be used for testing and validating interactions and responses of a vehicle controller within a simulated environment” and col. 3, ¶ 1: “The simulated vehicle can represent an autonomous vehicle that is controlled by an autonomous controller that can determine a trajectory, based at least in part, on the simulated environment”, e.g. col. 4, ¶¶ 3-4: “For purposes of illustration only, a simulated controller may represent different planning algorithms or driving behaviors ( e.g., different acceleration or braking profiled) that may introduce differences between the log data and simulated data…. In some instances, the SES can associate a controller with the simulated object, where the controller can determine trajectories for the object that deviate from the log data. Thus, the simulated object can perform actions associated with the object as represented in the log data and/or perform actions that deviate from the object as represented in the log data. Thus, the simulated object can intelligently react to deviations in the behavior of the simulated vehicle…”
to clarify, both the ego and non-ego vehicles of Kabirzadeh have a “simulated controller” associated with them, wherein each of their “simulated controller[s]” are reactive to other simulated elements, e.g. “In some instances, the simulated object models can 45 use controllers that allow the simulated objects to react to the simulated environment and other simulated objects (e.g., modeling physics-based behaviors and incorporating collision checking).” – e.g. “the simulated model applied to the simulated object 418 allows the simulation component 314 to determine that the simulated vehicle 416 has come to a stop [i.e. reactive to the ego vehicle] and prevent a collision by stopping at an earlier point than indicated by the log data 402…” in figure 4 wherein the ego vehicle was reactive to the pedestrian in the crosswalk in fig. 4 and wherein the non-ego vehicle (# 418) was reactive to the ego vehicle and note in col. 21, ln. 30-40: “In conducting the stop as depicted at time T 2B, the simulated vehicle 416 can perform a cost analysis. For example, the simulation component 314 can determine a cost associated with the simulated vehicle 416 applying a maximum braking (e.g., an emergency stop) which may result in excessive or uncomfortable accelerations for passengers represented in the simulated object 418.” - i.e. the ego vehicle # 416 was reactive to the non-ego vehicle # 418 in its stopping action, wherein the simulated object is then reactive so as to “prevent a collision by stopping at an earlier point” when it determines that the “simulated vehicle 416 has come to a stop”
While Kabirzadeh does not explicitly teach the following, Kabirzadeh in view of Albrecht teaches:
…probabilistically by: determining a set of goals or behaviours for the real-world agent; for each of the goals or behaviours, determining an expected trajectory model, wherein the expected trajectory model simulates future behaviour based on the goal or behaviour; and comparing the observed trace of the real-world agent with the expected trajectory model for each of the goals or behaviours, to determine a likelihood of each of the goals or behaviours, thus determining a distribution over the goals or behaviours, and extract a driving scenario defining a version of the road layout and at least one non-ego agent to be simulated, the non-ego agent associated with an inferred goal or behaviour for implementing in simulation within the defined version of the road layout;
Kabirzadeh, as was cited above, wherein this first determines trajectories from log data, and then determines waypoints for the trajectories as goals for the later simulated trajectories to reach
as taken in view of Albrecht, abstract, then see § I ¶¶ 3 and 5-7, then see § III including: “Our proposed system approximates the optimal policy as follows: For each other vehicle, enumerate its possible goals and inversely plan for that vehicle to each goal, giving the probabilities and predicted trajectories to the goals. The resulting goal probabilities and trajectories inform the simulation process of a Monte Carlo Tree Search (MCTS) algorithm to generate an optimal maneuver plan for the ego vehicle…” – then see the following subsections, including in particular see algorithm # 1 and § III.F (Examiner noting as well the subsections of section III as being relevant): “By assuming that each vehicle i 2 I seeks to reach one of a finite number of possible goal locations gi 2 Gi, using plans constructed from our defined macro actions, we can use the framework of rational inverse planning [4, 20] to compute a Bayesian posterior distribution over vehicle i’s goals at time t,… where L(s1:tjgi) is the likelihood of i’s observed trajectory given goal gi, and p(gi) specifies the prior probability of gi… The likelihood is a function of the reward difference between two plans: the reward ^r of the optimal trajectory from i’s initial observed state si 1 to goal gi after velocity smoothing, and the reward r of the trajectory which follows the observed trajectory until time t and then continues optimally to goal gi, with smoothing applied only to the trajectory after t… 1) Goal Generation: A heuristic function is used to generate the set of possible goals Gi for vehicle i based on its location and context information such as road layout and traffic rules. In our system, we include one goal for the end of the vehicle’s current road and goals for end of each reachable connecting road, bounded by the ego vehicle’s view region (as shown in Figure 1)… Inverse Planning: Inverse planning is done using A* search [12] over macro actions. A* starts after completing the current maneuver !i which produces the initial trajectory ^s1: . Each search node q corresponds to a state s 2 S, with initial node at state ^s , and macro actions are filtered by their applicability conditions applied to s. A* chooses the next macro action leading to a node q0 which has lowest estimated total cost1 to goal gi, given by f(q0) = l(q0) + h(q0). The cost l(q0) to reach node q0 is given by the driving time from i’s location in the initial search node to its location in q0, following the trajectories returned by the macro actions leading to q0… Our system predicts multiple plausible trajectories for a given vehicle and goal [example of a trajectory model which simulates future behavior based on the goal], rather than a single optimal trajectory. This is required because there are situations in which different trajectories may be (near)optimal but may lead to different predictions which could require different behaviour on the part of the ego vehicle. To predict multiple trajectories and associated probabilities to a given goal, we run A* search for a fixed amount of time and let it compute a set of plans with associated rewards (up to some fixed number of plans). Any time A* search finds a node that reaches the goal, the corresponding plan is added to the set of plans….” – see algorithm 2 as well to further clarify
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings from Kabirzadeh on which determined waypoints associated with a trajectory with the teachings from Alrecht on “an integrated planning and prediction system which leverages the computational benefit of using a finite space of maneuvers, and extend the approach to planning and prediction of sequences (plans) of maneuvers via rational inverse planning to recognise the goals of other vehicles”. The motivation to combine would have been that “Knowledge of the goals of other vehicles enables prediction of their future maneuvers and trajectories in relation to their goals. Such longterm
predictions facilitate planning over extended timescales to realise opportunities which might not otherwise present themselves, as illustrated in example in Figure 1.” (Albrecht, § I, second to last paragraph) and see § V: “We proposed an autonomous driving system which integrates planning and prediction over extended horizons, by leveraging the computational benefit of utilising a finite maneuver library…Our evaluation showed that the system robustly recognises the goals of other vehicles in diverse urban driving scenarios, resulting in improved decision making while allowing for intuitive interpretations of the predictions to justify (explain) the system’s decisions…”
Regarding Claim 2
Kabirzadeh teaches:
The computer system of claim 1, comprising a test oracle configured to evaluate the performance of the AV stack in the simulation, by receiving a simulated ego trace of the ego agent, as generated in the simulation, and scoring the simulated ego trace against a set of predetermined performance metrics.
Kabirzadeh, col. 18 last paragraph to col. 19 ¶ 3: “Additionally, the simulation component 252 can determine an outcome for the simulated scenario. For example, the simulation component 252 can execute the scenario for use in a simulation for testing and validation. The simulation component 252 generate the simulation data indicating how the autonomous controller performed (e.g., responded) and can compare the simulation data to a predetermined outcome and/or determine if any predetermined rules/assertions were broken/triggered. In some instances, the predetermined rules/assertions can be based on the simulated scenario (e.g., traffic rules regarding crosswalks can be enabled based on a crosswalk scenario or traffic rules regarding crossing a lane marker can be disabled for a stalled vehicle scenario). In some instances, the simulation component 252 can enable and disable rules/ assertions dynamically as the simulation progresses. For example, as a simulated object approaches a school zone, rules/assertions related to school zones can be enabled and disabled as the simulated object departs from the school zone. In some instances, the rules/assertions can include comfort metrics that relate to, for example, how quickly an object can accelerate given the simulated scenario… Successful validation of a proposed controller system may subsequently be downloaded by ( or otherwise transferred to) a vehicle for further vehicle control and operation.”
Then, see col. 21: “In conducting the stop as depicted at time T 2B, the simulated vehicle 416 can perform a cost analysis. For example, the simulation component 314 can determine a cost associated with the simulated vehicle 416 applying a maximum braking (e.g., an emergency stop) which may result in excessive or uncomfortable accelerations for passengers represented in the simulated object 418. Additionally, the simulation component 314 can determine a cost associated with the simulated vehicle 416 applying a minimum braking force, which may result in a collision or near-collision with the simulated object 418. The simulation component 314 can, using cost minimization algorithms (e.g., gradient descent), determine a trajectory to perform the stop in a safe manner while minimizing or optimizing costs.” – i.e. there was a plurality of predetermined “comfort metrics” such as ones associated with the acceleration of the simulated vehicle that the simulated was scored with for whether or not is passed/failed, wherein these were also in dependence upon the simulated non-ego trajectory (e.g. so as to “stop in a safe manner”)
Regarding Claim 5
Kabirzadeh teaches:
The computer system of claim 1, wherein the applying of the agent decision logic comprises matching target motion values along a spatial agent path, but with deviation from the target motion values permitted in reaction to the ego agent, to the spatial agent path associated with the inferred goal or behaviour. (Kabirzadeh, as was cited above for the applying limitation, e.g. col 4, lines 45-55: “For purposes of illustration only, in a scenario where a simulated object is following a simulated vehicle, if the simulated vehicle begins braking sooner or more aggressively than the braking behavior represented in the log data, the simulated object may modify its trajectory by braking sooner or changing lanes, for example, to avoid a collision or near-collision with the simulated vehicle.” – see above citations to clarify, i.e. the simulated object is following the simulated vehicle, simulated vehicle brakes (col. 21, ¶¶ 1-2), then the simulated object/non-ego vehicle brakes sooner to avoid a collision
Regarding Claim 9
Kabirzadeh in view of Albrecht teaches:
The computer system of claim 1, wherein the expected trajectory model is a single predicted trajectory associated with one of the each of the goals or behaviors or a distribution of predicted trajectories associated with the one of the each of the goals or behaviors. (Kabirzadeh, as was taken in view of Albrecht above)
Regarding Claim 10.
Kabirzadeh in view of Albrecht teaches:
The computer system of claim 1, wherein the one or more processors are configured to use the observed trace to predict a first trajectory model for the goal or behaviour and compare the first trajectory model with the expected trajectory model, wherein the first trajectory model has an observed portion between a first time and a second time that matches the observed trace. (Kabirzadeh, as was taken in view of Albrecht above; to clarify, see instant disclosure pages 49-50, subsection E, see Albrecht § III.E; see instant disclosure page 50-51 for the goal recognition including algorithm 1, see Albrecht § III.F and algorithm 1; see instant disclosure pages 52-53 subsection on inverse planning, see Albrecht § III.F, # 3 for inverse planning (including note footnote 1),
e.g. in § III.F: “The likelihood is a function of the reward difference between two plans: the reward ^r of the optimal trajectory from i’s initial observed state si 1 to goal gi after velocity smoothing, and the reward r of the trajectory which follows the observed trajectory until time t and then continues optimally to goal gi, with smoothing applied only to the trajectory after t. Then, the likelihood is defined as…”
Regarding Claim 11.
Kabirzadeh in view of Albrecht teaches:
The computer system of claim 10, wherein a defined reward function is applied to both the expected trajectory model and the first trajectory model for each goal, to compare respective rewards of the expected trajectory model and the first trajectory model. (Kabirzadeh, as was taken in view of Albrecht above)
Regarding Claim 13.
Kabirzadeh in view of Albrecht teaches:
The computer system of claim 1, wherein the one or more processors are configured to sample a first goal or behaviour from the distribution over a set of goals or behaviours, the agent decision logic configured to determine the closed- loop behaviour of the non-ego agent based on the sampled first goal or behaviour.. (Kabirzadeh for applying the agent decision logic to the non-ego agent/vehicle so it stops before colliding with the ego vehicle/reactive to what the ego vehicle does
as was taken in view of Albrecht, § III.G, ¶¶ 1- 2: “To compute an optimal plan for the ego vehicle, we use the goal probabilities and trajectory predictions to inform a Monte Carlo Tree Search (MCTS) algorithm [6]… The algorithm performs a number of simulations ^st:n, starting in the current state ^st = st down to some fixed search depth or until a goal state is reached. At the start of each simulation, for each other vehicle, we first sample a current maneuver, then goal, and then trajectory for the vehicle using the associated probabilities (cf. Section III-F).2 The sampled trajectories will be used to simulate the motion of the other vehicles, in closed-loop or open-loop modes (detailed below)”, see algorithm 2 to clarify, noting the title of the subsection is “Ego Vehicle Planning”
Regarding Claim 15.
Rejected under a similar rationale as claim 2 above.
Regarding Claim 18.
Kabirzadeh teaches:
The computer system of claim 1, wherein the agent decision logic is tuned so as to cause the non-ego agent to realize a trajectory that substantially corresponds to the observed trace in an event the behaviour of the ego agent in the simulation substantially matches the behaviour of a real ego agent in the real-world driving data. (Kabirzadeh, as cited above for the applying step, then see col. 2, last paragraph: “The simulated scenario can be identical to the captured environment or deviate from the captured environment” and col.5 ¶ 1: “Additionally, the log data can indicate that the vehicle attempted to perform a lane change action to the right but did not complete the action due to the object. In order to preserve this interaction in the simulated scenario, the SES can assign a waypoint associated with the trajectory of the simulated object such that the simulated object performs the lane change action and remains on the right side of the simulated vehicle Controllers associated with the simulated vehicle may then, in such examples, determine controls for the simulated vehicle which adapt to changes in the scenario while minimizing deviations from the paths as originally taken”
Regarding Claim 19.
Rejected under a similar rationale as claim 1 above.
Regarding Claim 20.
Rejected under a similar rationale as claim 2 above.
Regarding Claim 21.
Rejected under a similar rationale as claim 2 above, in particular note col. 19 ¶ 3: “Successful validation of a proposed controller system may subsequently be downloaded by ( or otherwise transferred to) a vehicle for further vehicle control and operation.” – in view of col. 4, ¶ 3: “In some instances, the simulated vehicle can be controlled using a controller that is the same as or that is different than the controller used by the vehicle that generated the log data… For purposes of illustration only, a simulated controller may represent different planning algorithms or driving behaviors ( e.g., different acceleration or braking profiled) that may
introduce differences between the log data and simulated data.” – and col. 3, ¶ 2, and col. 21, ¶ 1: “By way of example, if the motion of simulated object 418 relied solely on the log data 402, the simulated object 418 would have collided with the simulated vehicle 416. However, the simulated model applied to the simulated object 418 allows the simulation component 314 to determine that the simulated vehicle 416 has come to a stop and prevent a collision by stopping at an earlier point than indicated by the log data 402.”
Regarding Claim 24.
Rejected under a similar rationale as claim 1 above.
Claim(s) 3-4 and 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kabirzadeh, US 11,150,660 taken in view of Albrecht, Stefano V., et al. "Integrating planning and interpretable goal recognition for autonomous driving." arXiv preprint arXiv:2002.02277 (Feb. 6th, 2020) in view of Kakade, Hrishikesh, et al. "Autonomous Highway Overtaking." Han University of Applied Sciences.Master’s (2018).
Regarding Claim 3
While Kabirzadeh alone does not explicitly teach the following, Kabirzadeh in view of Kakade teaches:
The computer system of claim 2, wherein the one or more processors are configured to provide an output comprising a score-time plot for each performance metric. (Kabirzadeh as cited above for claim 2 incl. col. 2 ¶ 2 incl.: “…the rules/assertions can include comfort metrics that relate to, for example, how quickly an object can accelerate given the simulated scenario.”
Taken in view of Kakade, abstract last two paragraphs, in particular: “The scope of this thesis is to develop the test protocol and automated driving system for the accelerative/normal and flying overtaking maneuver in a highway context. Firstly, the test protocol for the overtaking maneuver is developed which is in accordance with the ISO standards (ACC, LCA, BSM, LKA, and LDW), Euro NCAP (safety assist protocols), and various rules and regulations set by few governments (The Netherlands, UK, and Province of Alberta) for overtaking maneuver… Finally, the developed test protocol is verified and validated analytically (using mathematical equations) as well as virtually (automated driving system using PreScan) to check if the host vehicle performs the overtaking maneuver fail-safely… The sensor modeling, ADAS implementation, and the development of the complete automated driving system for the overtaking maneuver is carried out using PreScan and MATLAB/Simulink software.”
Then see § 7.1 for the “Scenario evaluation”, in particular in “Use case 1” in § 7.1.1 see the subsection on “Simulation results” incl. # 1 and # 2, incl.: “The speed profile of the host vehicle during the overtaking maneuver is shown in the MATLAB figure 7.7. Host vehicle speed starts decreasing with ACC at t = 22.3 s and becomes 100 km/h at t = 24.5 s. At this point, the overtaking headway threshold condition is reached and the left lane change phase starts while the speed of host vehicle also starts increasing gradually. Host vehicle achieves its initial set speed at the end of the passing phase (t = 34.8 s).” and # 3: “The acceleration profile of the host vehicle during the overtaking maneuver is shown in the MATLAB figure 7.8. The maximum deceleration and acceleration during the maneuver is limited to -3.1 and 1.8 m/s2 respectively which is within the permissible limits as discussed in section 4.8.” – and see fig. 7-8, i.e. this is an example of a score-time plot, wherein this is within the thresholds specified in § 4.8; see § 7.1.2 # 2-3 as well- see the figures in particular
To clarify, see § 4.8, incl. # 3: “The maximum lateral acceleration, lateral deceleration, and axial acceleration was limited to 1.1, 0.41, 2.5 m/s2 respectively in [48]. As per the ISO standard 15622 (ACC), the average automatic deceleration of ACC systems shall not exceed 3.5 m/s2 (average over 2 s), the average rate of change of automatic deceleration (negative jerk) shall not exceed 2.5 m/s3 (average over 1 s). Automatic acceleration of ACC systems shall not exceed 2 m/s2 [67]. Thus, the longitudinal acceleration ranges between 2.5 and 7 m/s2 while lateral acceleration ranges between 1.1 and 2.9m/s2 from the above investigation. Hence, the maximum longitudinal acceleration of 3.5 m/s2 is considered from the comfort aspects while developing test protocol for the highway overtaking maneuver.”
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings from Kabirzadeh as cited above including for claim 2 incl. col. 2 ¶ 2 incl.: “…the rules/assertions can include comfort metrics that relate to, for example, how quickly an object can accelerate given the simulated scenario.” with the teachings from Kakade on a “test protocol and automated driving system for the accelerative/normal and flying overtaking maneuver in a highway context” (Kakade, abstract) The motivation to combine would have been that “As per the ISO standard 15622 (ACC), the average automatic deceleration of ACC systems shall not exceed 3.5 m/s2 (average over 2 s), the average rate of change of automatic deceleration (negative jerk) shall not exceed 2.5 m/s3 (average over 1 s). Automatic acceleration of ACC systems shall not exceed 2 m/s2 [67]….the maximum longitudinal acceleration of 3.5 m/s2 is considered from the comfort aspects while developing test protocol for the highway overtaking maneuver."” – also, see abstract as cited above to clarify.
Regarding Claim 4
Rejected under a similar rationale as claim 3 above.
Regarding Claim 14.
Rejected under a similar rationale as claim 3 above.
Claim(s) 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kabirzadeh, US 11,150,660 taken in view of Albrecht, Stefano V., et al. "Integrating planning and interpretable goal recognition for autonomous driving." arXiv preprint arXiv:2002.02277 (Feb. 6th, 2020) in view of Wu, Mo, and Benjamin Coifman. "Quantifying what goes unseen in instrumented and autonomous vehicle perception sensor data–A case study." Transportation research part C: emerging technologies 107 (2019): 105-119.
Regarding Claim 17.
While Kabirzadeh in view of Albrecht does not explicitly teach the following, Kabirzadeh, in view of Albrecht and Wu, teaches:
The computer system of claim 1, wherein the one or more processors are configured to apply one or more offline perception algorithms to the real-world driving data, in order to extract the observed trace, wherein the offline perception algorithm is applied outside of runtime of the AV stack. (To clarify on the BRI, see page 20-21, paragraph split between the pages, also page 13 last paragraph
See Kabirzadeh, as was cited above for the waypoints based on the trajectory which was based on sensor data [real-world driving data] but this does not teach applying a non-real time perception algorithm to the sensor data for extracting the trajectory, wherein this is outside of the runtime of the AV stack
as taken in view of Wu, abstract including: “This paper fuses individual vehicle actuations from loop detectors and concurrent ambient vehicle trajectories collected from an IPV to provide a rare opportunity to assess the perception sensor performance in situ, using the loop detector data to “see” what is missed by the perception sensors… While the specific results are unique to the IPV, one should expect similar biases from other perception sensor data; though without an independent measure like the loop detectors in this study, it will be difficult to quantify what goes unseen in those data sets….” – see § 4 to clarify, and see § 2 for more details, in particular see §§ 2.1.2 and 2.1.2, then see § 2.1 including: “When the two data sets are synchronized and fused together they provide a much more complete picture than either on their own (e.g., the arrows in Fig. 1(d) show how disturbances passing the IPV can be matched to their passages at the bounding loop detector stations). The following subsections discuss the two data sources in detail”
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings from Kabirzadeh on extracting the waypoints based on the sensor-data with the teachings from Wu on fusing sensor data from vehicles with loop detectors. The motivation to combine would have been that “Perception sensors on instrumented probe vehicles (IPV's) are increasingly being used to collect data for traffic flow and driver behavior modeling, while the rapidly advancing field of autonomous vehicles (AV) uses perception sensors to identify and mitigate hazards. Yet, the perception sensors are far from perfect, with targets going undetected. The “unseen” targets might lead to inaccurate traffic flow theories from IPV data, or crashes involving AV's. This paper fuses individual vehicle actuations from loop detectors and concurrent ambient vehicle trajectories collected from an IPV passing over the detectors to provide a rare opportunity to assess the perception sensor performance, using the loop detector data to see what is missed by the IPV perception sensors. While the focus of this work is IPV data for offline study, the general findings are also relevant for perception sensors used in AV applications. Ultimately though, the data fusion work in this paper arises from the pursuit of high resolution empirical traffic data to better understand the nuances of traffic dynamics….” (Wu, § 4)
Claim(s) 22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kabirzadeh, US 11,150,660 taken in view of Albrecht, Stefano V., et al. "Integrating planning and interpretable goal recognition for autonomous driving." arXiv preprint arXiv:2002.02277 (Feb. 6th, 2020) in view of in view of Caldwell et al., US 11,565,709.
Regarding Claim 22.
Kabirzadeh in view of Caldwell teaches:
The method of claim 19, wherein the AV stack comprises at least one trainable machine learning component, the method performed multiple times as part of a structured training process, wherein the performance of the AV stack is evaluated in each simulation, and that evaluation is used to train parameters of the machine learning component.(Kabirzadeh, as cited above for claim 2, then see Caldwell, abstract: “…The computing system may calculate performance metrics associated with the actions performed by the vehicle in the simulation as directed by the autonomous controller. The computing system may utilize the performance metrics to verify parameters of the autonomous controller (e.g., validate the autonomous controller) and/or to train the autonomous controller utilizing machine learning techniques to bias toward preferred actions.
Then see Caldwell, fig. 3, for the “Metrics” boxes, and do note the “Comfort metric”, along with “Safety” and “Time to Destination” – clarified on in col. 4, ¶ 2 and col. 13, ¶¶ 2-4, and col. 20, ¶¶ 2-3 Then see col. 31, second to last paragraph: “Based on a determination that the performance metric is associated with the threshold level ("Yes" at operation 510), the process, at operation 512, the process may include training (e.g., modifying) the autonomous controller to bias 40 toward the vehicle action. In various examples, training the autonomous controller may include modifying one or more parameters associated therewith. In some examples, the training may be performed based on a cost associated with a vehicle action. In such examples, the computing system 45 may train autonomous controller to bias toward low cost vehicle actions. In various examples, one or more other vehicles may be controlled based on user input, such as via a user interface. In such examples, the training of the autonomous controller may be performed utilizing user 50 input. In some examples, the training may be performed utilizing machine learning techniques. In such examples, the data associated with the scenario (e.g., object position, object movement, vehicle position, vehicle action, etc.) may be utilized as training data to bias the autonomous controller 55 to perform the action in a similar scenario. In at least some examples, such machine learning techniques may comprise reinforcement learning.”
As to the doing multiple times, see the flowchart in fig. 5, i.e. it does this loop iteratively until the simulation results are evaluated to be successful (the “Yes” portions of fig. 5, see accompanying descriptions to clarify)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings from Kabirzadeh on the AV simulation system from Zoox, Inc. which had comfort metrics it scored simulations with with the teachings from Caldwell on a similar AV simulation system from Zoox, Inc. wherein this includes “The computing system may calculate performance metrics associated with the actions performed by the vehicle in the simulation as directed by the autonomous controller. The computing system may utilize the performance metrics to verify parameters of the autonomous controller (e.g., validate the autonomous controller) and/or to train the autonomous controller utilizing machine learning techniques to bias toward preferred actions.” (Caldwell, abstract). The motivation to combine would have been that “Thus, the techniques described herein may significantly improve the performance of the autonomous controller and greatly improve the safety of vehicle operations” (Caldwell, col. 2, ¶ 1).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DAVID A. HOPKINS whose telephone number is (571)272-0537. The examiner can normally be reached Monday to Friday, 10AM to 7 PM EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Ryan Pitaro can be reached at (571) 272-4071. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/David A Hopkins/Primary Examiner, Art Unit 2188
Prosecution Insights