DETAILED ACTION
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 .
Status of the Claims
Claims 1-19 were previously pending and subject to a non-final office action mailed 02/20/2026. Claims 1-12, 15-16, and 18-19 were amended; claims 13-14 and 17 were cancelled, and no claim was added in a reply filed 05/18/2026. Therefore claims 1-12, 15-16, and 18-19 are currently pending and subject to the final office action below.
Response to Arguments
Applicant’s arguments, see p. 8-15, filed 05/18/2026, with respect to 101 rejection have been fully considered and are persuasive. The 101 rejection of claims 1-19 has been withdrawn.
Applicant’s arguments with respect to 103 rejection have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
No arguments were presented in regards to 112(f) interpretation of claim 19. As such, the 112(f) interpretation is replicated here.
Claim Rejections - 35 USC § 103
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1, 3-7, 10-12, 15-16, 18-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Levis (US 2006/0235739) in view of Kumar (US 2008/0201019) in further view of NASA, “Project Planning and control handbook”, published by NASA on September 25, 2017, hereinafter “Nasa”, in further view of Borhan (US 2020/0398859).
As per claim 1/18-19, Levis discloses A method for controlling a plurality of vehicles performing missions along a respective route, the method comprising (paragraphs 3, 35, 46):
wherein the missions that the plurality of vehicles perform form a present set of missions, the method comprising performing a previous set of missions ([0088] For example, the process 29 may access a file containing Historical Data 28. Historical Data is reference data that can be used as an aid in determining whether and how to update the Original Dispatch Plan… [0090] This aspect of the Historical Data captures, in part, the "experience" aspect of a driver by way of storing past delivery information that is used to provide a benchmark to determine whether the execution of a Dispatch plan is on schedule or behind schedule.” The presently required deliveries and pickups in the current dispatch plan constitute the claimed present set of missions. Levis also stores and uses historical information from earlier dispatch plans and past workdays. Levis explains that historical data may indicate completed deliveries or pickups and that a service stop completion flag may be recorded for each stop. Levis further teaches that historical data may comprise ahistorical running average of time and location information associated with delivery stops. It explains that the current time, current location, and degree of completion of required tasks are compared with past average time and location measurements. Thus, the service stops performed during an earlier dispatch plan or prior workday constitute a previous set of missions, while the service stops presently being performed constitute a present set of missions).
Regarding “after an initial of the missions has started, and before all missions are completed, determining a mission completion deviation comprising a deviation of an actual number of completed missions from the desired number of completed missions,”
Levis teaches the temporal condition, the determination of completed missions, the actual completed mission count, and the comparison of actual completion progress with expected completion progress. Nasa, as discussed below, supplies the more specific quantitative comparison between the cumulative actual number of completed missions and the cumulative desired number of missions scheduled for completion.
Regarding, “after an initial of the missions has started and before all missions are completed”, Levis discloses
After execution of the present service stop mission set has begun and while service stop missions remain uncompleted, determining the number of proportion of service stop missions that have been completed ([0088] … Because deliveries that have been already completed by the driver are not be impacted by subsequent developments, such as weather or traffic, it is only the remaining deliveries in the Dispatch Plan that must be analyzed in order to produce an Updated Dispatch Plan. The fact that FIG. 3 illustrates Historical Data as separate from the Manifest Data is for conceptual purposes only and is not intended to limit how the Historical Data is stored. In some embodiments, the indication of which deliveries/pickups have been completed are stored in conjunction with the Manifest Data or Dispatch Plan. Thus, conceptually, this portion of the Historical Data could be viewed as an augmentation of the Dispatch Plan. Typically, a service stop completion flag in the Dispatch Plan is recorded indicating the service stop has been completed. Regardless of how the indication is recorded, data indicating past deliveries can be modeled as Historical Data… [125]… The level of completion ("completion status") of the Dispatch Plan can be easily determined by comparing the ratio of service stops completed (or packages delivered) with the total number of service stops (or packages). The previously mentioned completion flag or indicator provides an indication of whether the corresponding service stop in the record has been performed. Thus, completion of 30 service stops from a total of 120 represents 30/120=0.25 or 25% completion. This would correspond to point 304 on the line.” Based on the paragraphs, Levis teaches that already completed stops are separated from the remaining stops when updating the dispatch plan. It also discloses that the system makes its completion d3etermination after execution has begun and before all missions have been completed because the system identifies both completed service stops and service stops that remain to be performed. Thus, Levis teaches determining the actual completed mission component used in the claimed mission completion deviation.);
After execution has begun and before all missions have been completed, determining whether the actual completion of the service-stop missions deviates from the expected completion of the missions at the current time (“[0126] Another line 310 in FIG. 14 represents the time allocated for performing the work. Typically, this is a workday, with a defined number of hours (e.g., eight hours) … [0128] It is evident that these three of these metrics are interrelated and a mapping can occur from each of these metrics. The portable computer can track time, its location, and record the completion of a service stop in the Dispatch Plan. The portable computer can then compare the relative completion status of each metric.” Levis establishes an allocated work time continuum and compares that continuum with the actual proportion of completed service stop. It explains that 25 percent of an eight-hour workday corresponds to a 10am and that the portable computer compares the relative completion status of the work with the relative completion status of the allocated time. Levis thereby determines whether actual mission completion is ahead of, behind, or consistent with the expected schedule progress. Levis supplies the actual completion and schedule comparison portions of the claimed mission completion deviation determination. Nasa, as discussed below, supplies the more specific cumulative actual number versus desired number framework).
Using the identified mission completion schedule deviation as a basis for changing the subsequent operation of the vehicles during the remaining missions (“[0131] If the system determines that deliveries are "behind schedule", it can then trigger the Dispatch Updating Process to determine if the records in the Dispatch Plan can be sequenced more efficiently or simply notify the user appropriately…” Levis teaches notifying central dispatch so that additional resources may be allocated. This teaching does not itself disclose determining a different second balance parameter value. Rather, it establishes that a mission completion deviation is used to trigger a change in the operation of the remaining missions. Kumar supplies the particular cost/progress balance adjustment claimed).
However, Levis does not disclose but Kumar discloses the first balance parameter, establishing mission segment completion timing in dependence on that balance, revising the cost/progress balance in response to schedule deviation, determining a velocity profile for the remaining route, and controlling the propulsion arrangement according to that profile.
obtaining for each vehicle of the plurality of vehicles, a first value of a balance parameter, indicative of a balance between a cost for operating the respective vehicle along at least a part of the route, and a progress of the respective vehicle along at least a part of the route (paragraph 56, 75, 77-79, 88” The coefficients of the linear combination depend on the importance (weight) given to each of the terms. Note that in equation (OP), u(t) is the optimizing variable that is the continuous notch position. If discrete notch is required, e.g. for older locomotives, the solution to equation (OP) is discretized, which may result in lower fuel savings. Finding a minimum time solution (.alpha..sub.1 set to zero and .alpha..sub.2 set to zero or a relatively small value) is used to find a lower bound for the achievable travel time (T.sub.f=T.sub.fmin). In this case, both u(t) and T.sub.f are optimizing variables. “, “ [0075] Throughout the document exemplary equations and objective functions are presented for minimizing locomotive fuel consumption. These equations and functions are for illustration only as other equations and objective functions can be employed to optimize fuel consumption or to optimize other locomotive/train operating parameters” Kumar teaches optimizing a vehicle trip according to an objective function containing weighted terms for operating cost and vehicle progress. Kumar identifies minimization of total fuel consumption and minimization of travel time as optimization objectives. Kumar’s fuel consumption, emissions, braking loss, and equipment operation represent costs of operating the vehicle. Travel time, required arrival time, distance traveled, and progress toward the destination represent progress of the vehicle along the route. The coefficients or weights assigned to those respective terms therefore constitute a balance parameter indicative of the relative importance assigned to operating cost and progress. Kumar further teaches that the tradeoff is established for respective vehicles. “[0088]… For example, the travel-time fuel use tradeoff curve as illustrated in FIG. 4 reflects a capability of a train on a particular route at a current time, updated from ensemble averages collected for many similar trains on the same route. Thus, Kumar teaches obtaining, for each vehicle, a first value of a parameter establishing the relative balance between vehicle operating cost and vehicle progress “.)
In regards to “characterized by - establishing, in dependence on the first balance parameter values, a desired number of completed missions as a function of time”, Kumar discloses establishing planned completion times for respective mission specific vehicle segments in dependence on the first cost/progress balance-parameter values ([0104] In an exemplary embodiment, the present invention is able to break down a longer trip into smaller segments in a special systematic way. Each segment can be somewhat arbitrary in length, but is typically picked at a natural location such as a stop or significant speed restriction, or at key mileposts that define junctions with other routes. Given a partition, or segment, selected in this way, a driving profile is created for each segment of track as a function of travel time taken as an independent variable, such as shown in FIG. 4. The fuel used/travel-time tradeoff associated with each segment can be computed prior to the train 31 reaching that segment of track….[0105] FIG. 4 depicts an exemplary embodiment of a fuel-use/travel time curve. As mentioned previously, such a curve 50 is created when calculating an optimal trip profile for various travel times for each segment. That is, for a given travel time 49, fuel used 53 is the result of a detailed driving profile computed as described above. Once travel times for each segment are allocated, a power/speed plan is determined for each segment from the previously computed solutions” Kumar distributes the available travel time among the segments so that the required total trip time is satisfied while total fuel consumption is minimized. Thus, based on the above, Kumar establishes when successive mission specific segments or stops are intended to be completed, and those completion times depend on the selected fuel versus travel time balance. Nasa supplies the act of expressing such scheduled completion times as a cumulative desired number of completed missions as a function of time).
determining for each of one or more of the plurality of vehicles a second balance parameter value, different from the respective first balance parameter value, the respective second balance parameter value being dependent on the mission completion deviation (“[0017] In another exemplary embodiment a method discloses providing an optimized mission plan to be manually applied. The optimized mission plan is re-planned in response to a manual mission plan being implemented. The manual plan is adjusted when the manual plan deviates from the optimized plan by more than a predetermined amount.”, “[0091] … Using the actual speed, power and location of the locomotive, a comparison is made between a planned arrival time and the currently estimated (predicted) arrival time 25. Based on a difference in the times, as well as the difference in parameters (detected or changed by dispatch or the operator), the plan is adjusted 26.” “[0092] A re-plan may also be made when it is desired to change the original objectives. Such re-planning can be done at either fixed preplanned times, manually at the discretion of the operator or dispatcher, or autonomously when predefined limits, such a train operating limits, are exceeded. For example, if the current plan execution is running late by more than a specified threshold, such as thirty minutes, the exemplary embodiment of the present invention can re-plan the trip to accommodate the delay at expense of increased fuel as described above or to alert the operator and dispatcher how much of the time can be made up at all (i.e. what minimum time to go or the maximum fuel that can be saved within a time constraint).” “[0147] … This information (actual estimated arrival time or information needed to derive off-board) can also be communicated to the dispatch center to allow the dispatcher or dispatch system to adjust the target arrival times. This allows the system to quickly adjust and optimize for the appropriate target function (for example trading off speed and fuel usage). “, “[0177] FIG. 20 depicts another closed loop system where an operator is in the loop. The optimizer 650 generates the power/operating characteristic required for the optimum performance. The information is communicated to the operator 647, such as but not limited to, through human machine interface (HMI) and/or display 649. This could be in various forms including audio, text or plots or video displays. The operator 647 in this case can operate the master controller or pedals or any other actuator 651 to follow the optimum power level. [0178] If the operator follows the plan, the optimizer continuously displays the next operation required. If the operator does not follow the plan, the optimizer may recalculate/re-optimize the plan, depending on the deviation and the duration of the deviation of power, speed, position, emission etc. from the plan. If the operator fails to meet an optimize plan to an extent where re-optimizing the plan is not possible or where safety criteria has been or may be exceeded, in an exemplary embodiment the optimizer may take control of the vehicle to insure optimize operation, annunciate a need to consider the optimized mission plan, or simply record it for future analysis and/or use. In such an embodiment, the operator could retake control by manually disengaging the optimizer.” Based on the above, Kumar teaches determining a revised optimized mission plan in response to a deviation of actual vehicle operation from the original optimized mission plan. Paragraph 17 establishes that a value or operating state governing the revised plan differs from the value or operating state governing the original plan and that the revision is caused by a deviation. Based on paragraph 147, Kumar discloses obtaining a revised balance value that changes the tradeoff between speed or schedule progress and fuel or operating cost relative to the balance governing the original optimized plan. The “target function” is the objective function whose terms are assigned the coefficient or weights described in paragraph 78. Adjusting and reoptimizing that target function changes the importance assigned to fuel usage relative to speed or target arrival time and therefore obtains a balance parameter value different from the original first balance parameter value. Paragraph 9192 confirms that the revised plan uses a different cost/progress balance: the original plan places relatively greater emphasis on fuel conservation, while the revised plan accepts increased fuel cost to recover schedule time. Levis and Nasa supply the particular claimed mission completion deviation. Using the deviation between the actual and desired completed mission counts determined in the Levis-Nasa system as Kumar’s detected departure from the optimized mission plan causes Kumar to recalculate the target function and obtain a second speed/fuel balance parameter value that differs from the first value and depends on the mission completion deviation).
determining, in dependence on the respective second balance parameter value, a respective velocity profile for a respective of the one or more of the plurality of vehicles for at least a portion of the respective remainder of the respective route ([105] … If the locomotive consist or train changes significantly along the route, e.g. from loss of a locomotive or pickup or set-out of cars, then driving profiles for all subsequent segments must be recomputed creating new instances of the curve 50. These new curves 50 would then be used along with new schedule objectives to plan the remaining trip…[0106] Once a trip plan is created as discussed above, a trajectory of speed and power versus distance is used to reach a destination with minimum fuel and/or emissions at the required trip time…[0119] After completing a re-plan from the collection of events described above, the new optimal notch/speed plan can be followed using the closed loop control described herein.” The speed versus distance trajectory constitutes the claimed velocity profile. Because it is generated from the adjusted fuel versus travel time target function. It is determined in dependence on the second balance parameter value. Because the subsequent segments are recomputed after the trip begins, it applies to at least a portion of the respective remainder of the route.)
controlling a propulsion arrangement of the respective of the one or more of the plurality of vehicles according to the respective determined velocity profile ([106]…In another exemplary embodiment of the present invention commands for powering and braking are provided as required to follow the desired speed-distance path…[107]… Feedback control strategies compare the actual speed as a function of position to the speed in the desired optimal profile. Based on this difference, a correction to the optimal power profile is added to drive the actual velocity toward the optimal profile…[0144] Using exemplary embodiments of the present invention, the train may operate in a plurality of operations. In one operational concept, an exemplary embodiment of the present invention may provide commands for commanding propulsion, dynamic braking. “).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of the invention to include the limitation above as taught by Kumar in the teaching of Levis, in order to optimize fuel efficiency, emission output, vehicle performance, infrastructure and environment mission performance of the diesel powered system (please see Kumar paragraph 4).
However, Levis in view of Kumar does not disclose but Nasa discloses
Establishing, according to a time phased baseline schedule, a desired cumulative number of completed tasks as a function of time (“Conducting on-going trend analysis on the schedule baseline plan vs. actual completions is a valuable technique to use when monitoring schedule performance trends. As shown in Figure 62, this metric reflects the basic monthly cumulative total of tasks/milestones that have actually been completed to-date versus the cumulative total of schedule items that should be completed to-date per the baseline plan.” The cumulative total of schedule items that should be completed by successive times is a desired number of completed tasks as a function of time. When the Nasa framework is applied to Levis’s service stop missions, each service stop is a discrete task, and the cumulative number of service stops scheduled for completion by each time is the desired number of completed missions as a function of time. Kumar establishes the planned completion times based on its first cost/progress balance values. Nasa establishes the cumulative scheduled count from those completion times. Constructing Nasa’s cumulative sch3eduled task baseline using the segment or service stop completion times establish according to Kumar’s first balance parameter values results in establishing the desired number of completed missions as a function of time in dependence on the first balance parameter values.)
During execution of a scheduled set of tasks, determining a quantitative schedule completion departure by comparing the cumulative actual number of completed tasks with the cumulative number scheduled for completion at the same time (fig. 65, “BEI = cumulative number of baseline tasks completed divided by cumulative number of baseline tasks scheduled for completion.” Nasa explains that a value greater than 1.0 indicates performance ahead of the baseline and progressively lower values indicate poorer completion performance. Nasa also defines schedule variance as “Budgeted Cost of Work Performed (BCWP) – Budgeted Cost of Work Scheduled (BCWS) SV can also be stated as the value of the work completed minus the value of the work that was planned to be completed.” The BEI comparison directly uses: the actual cumulative number of completed tasks; and the cumulative number scheduled for completion. Levis supplies the temporal context because it performs the comparison during execution, after some stops have been completed and while other stops remain. Applying Nasa’s cumulative actual versus scheduled completion comparison during Levis’s ongoing dispatch plan execution results in determining after an initial mission has started and before all missions are completed, the claimed mission completion deviation”. )
In regards to “a deviation of an actual time to complete all missions in the previous set of missions, from a desired time to complete all missions in the previous set of missions”, Nasa discloses Determining, for a completed prior set of scheduled tasks, a deviation between an actual finish time and a corresponding desired baseline finish time (Nasa teaches retaining baseline and actual task finish dates and comparing actual completion with baseline completion. Nasa states that tasks and milestones should be updated to reflect “As project work is executed, all tasks/milestones in the schedule should be updated to reflect their current status. This will involve timely updates to network logic, task percent completes, resource allocations, remaining durations, and actual start and finish dates.” Nasa gives the example “In this example, Task 1 was completed at the end of Q4, approximately 4 ½ months later than the baseline plan.” When the above is applied to a completed Levis dispatch plan: the baseline finish time of the final service stop mission is the desired time to complete all missions in the previous set; the actual finish time of the final service stop mission is the actual time to complete all missions in that set; and their difference is the claimed total completion time deviation”).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of the invention to apply the time phased task completion framework as taught by Nasa to the delivery mission management system of Levis, in order to express how well the project is following the baseline plan (Nasa, page 130)
However, the combination of Levis in view of Kumar and Nasa does not disclose but Borhan regarding “determining, for the previous set of missions, a value of a reward parameter, in dependence of a deviation of an actual time to complete all missions in the previous set of missions” discloses “determining a reward or penalty value in dependence on a comparison between observed vehicle performance and a desired performance condition” ([0003] Reinforcement learning is an area of machine learning. Reinforcement learning involves setting a goal for a system or agent, then allowing the agent to implement and update a policy to meet the goal. The agent implements the policy and in turn receives a reward when successful and no reward or a penalty when unsuccessful.…[0016]…The observations are analyzed by the controller (for example, via comparison to a defined goal or a stored database of results) and a reward is received when a desirable result is achieved in the environment. The controller updates the base policy each time an observation is made so that the policy trends toward desirable results and the policy is updated in view of the results observed in the environment….[0020]… Based on the observed vehicle operational parameters, a reward or penalty is issued. In some embodiments, the reward or penalty is issued based on a threshold condition (e.g., a fuel economy above or below a threshold, an engine speed overshoot or undershoot, a gear shift timing, a NOx conversion efficiency, etc.). In some embodiments, the reward or penalty is issued based on look-up tables or matrixes of desirable results in various conditions. In some embodiments, the reward or penalty is issued based on one or more algorithms designed to optimize a set of parameters (e.g., speed control and fuel economy, NOx conversion efficiency and ammonia slip, etc.) ….[0022]… At step 62, a reward or penalty is issued based on the observed state from step 58. In some embodiments, determining the reward or penalty includes querying a remote database of stored outcomes…. [0032]… The policy circuit 86 is also structured to receive reward or penalty signals (e.g., values, data, information, messages, etc.) from the reward/policy circuit 98 and update the policy based on the received reward or penalty signals.” Based on the above, Borhan teaches establishing a desired goal, evaluating whether vehicle operation successfully achieves that goal, and providing a reward or penalty based on the result. It also teaches that the reward or penalty may be based on the observed vehicle operating parameters and may be determined using a threshold, matrix, lookup table, or optimization algorithm. The speed control and fuel economy example is relevant because Kumar’s balance parameter similarly governs a tradeoff between vehicle progress and operating cost. In this case, Nasa supplies the particular observed result and desired goal required by the claim. As explained in the Nasa section, Nasa determines the deviation between the actual finish time and the desired baseline finish time for a completed prior mission set. Using Nasa’s deviation between the actual and desired times to complete a previous Levis mission set as Borhan’s comparison between the observed vehicle operation result and the defined goal results in determining Borhan’s reward parameter value in dependence on the claimed completion time deviation. Borhan does not independently disclose that the reward is specifically based on the total completion time deviation of a mission set. Levis and Nasa supply that specific performance condition, Borhan supplies determining the reward value from the comparison with that condition).
Regarding “wherein the respective second balance parameter value is determined in dependence on the previous set of missions”, Borhan discloses “updating a vehicle control parameter in dependence on a reward or penalty value” ([0023] The updating of the policy at step 66 includes updating control parameters used for control of the vehicle system 22. For example, the policy may be a PID control scheme and the update to the policy may include updating one or more parameters of the PID. In some embodiments, the policy includes a neural network or a deep neural network and one or more levels of the neural network is adjusted or updated in view of the reward or penalty… [0020]… The policy can be stored locally on the vehicle controller 30 and updated based on the feedback reward or penalty queried from the remote controller 34. Once the policy is updated in view of the reward or penalty, the vehicle controller 30 controls the vehicle system 22 based on the updated policy and the cycle of reinforcement learning continues.” Borhan expressly teaches that the reward or penalty does not merely classify the prior result as desirable or undesirable. The reward or penalty causes the machine learning policy to update the control parameters used to operate the vehicle. Kumar’s coefficients or weights establish the relative importance assigned to speed or mission progress and fueled or operating cost. Those coefficients are vehicle control and trip optimization parameters within the meaning of Borhan paragraph 23. Moreover, Borhan expressly contemplates using its reward process to optimize the same types of parameters that Kumar balances “speed control and fuel economy”. Thus, including Kumar’s speed versus fuel weighting coefficient among the vehicle control parameters updated by Borhan’s policy causes Kumar’s second balance parameter value to be determined in dependence on Borhan’s reward parameter value”)
Regarding “wherein the respective second balance parameter value is determined in dependence on an outcome of a machine learning process at the previous set of missions, the machine learning process being dependent on the reward parameter values determined for the respective earlier previous set of missions”, Borhan discloses “determining an updated vehicle control policy as the outcome of a repeated machine learning process that depends on reward or penalty values generated during earlier vehicle operation iterations” ([0003]… With repeated iterations of rewards and/or penalties, the policy is updated to achieve a maximum incidence of reward. n this way, the agent continues to improve the policy in an ongoing manner over time.…[0042]… The resulting reward signals and penalty signals can be used to develop a neural network, a deep neural network, or other networks and/or learning policy schemes that replace the base policy over time. n this way, the policy is continually updated and improved to control operation of the DEF doser 114 and improve operation of the aftertreatment system 106 based on real world information. “ the policy produced during a later iteration is not based solely on the reward generated in that iteration. It incorporates the effects of the rewards and penalties generated during earlier iterations. Levis supplies the successive sets of missions performed during successive dispatch periods. Borhan supplies the cyclic learning procedure in which each earlier operating iteration generates a reward value that contributes to the policy used during later iterations. Performing Borhan’s cyclic reward and policy update during successive Levis dispatch plan mission sets causes the policy produced at the claimed previous set of missions to depend on the reward parameter values generated for the respective earlier previous sets of missions. The updated policy produced during the previous mission set is therefore an outcome of a machine learning process that is dependent on rewards from the earlier previous mission sets. “[0016] … The use of machine learning allows the vehicle system to update the control policy based on received real world results or based on the particular arrangement of a vehicle. In this way, initial tuning can be improved and ongoing system control can account for changing operating conditions of components within the vehicle system… [0023] The updating of the policy at step 66 includes updating control parameters used for control of the vehicle system 22… [0039]… Reinforcement learning control processes within the vehicle controller 30 utilize online or real-time optimization with data collected in real conditions to calibrate and optimize the parameters which can adapt to the change in operating condition without any need to re-calibrate manually. Additionally, when the vehicle system 22 runs in a non-calibration condition (e.g., normal operation), the vehicle controller 30 can continue to learn and/or calibrate relevant parameters and the updated policy due to the adaptive nature of reinforcement learning algorithms.” Kumar supplies the particular control parameter claimed – the coefficient or weight balancing fuel cost and mission progress. Using the updated Borhan policy to select or update Kumar’s speed versus weighting causes the second balance parameter value to be determined in dependence on the outcome of Borhan’s machine learning process. Thus, the second balance parameter value is determined both in dependence on the reward value generated for the previous mission set; and in dependence on the cumulative policy outcome resulting from reward values generated during the earlier previous mission sets. Borhan also discloses that the remote controller can compute updated policy information for other vehicles using reward and sensor information (paragraph 25, 33). In claim 15, independently confirms this multi vehicle arrangement by reciting first and second vehicle systems. Thus, Borhan’s machine learning process is applicable to “each of one or more of the plurality of vehicles” managed by Levis).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of the invention to apply the reward and policy update as taught by Borhan to the combination of Levis in view of Kumar and Nasa in order to calibrate and optimize the parameters which can adapt to the change in operating condition without any need to re-calibrate manually (paragraph 39, Borhan)(and in order to use Nasa’s previous mission set completion time deviation as an observed performance result; determine a reward or penalty based on whether that result satisfies the desired completion time goal, update a machine learning policy using that reward and the rewards generated during earlier mission sets; use the updated policy to determined Kumar’s revised speed versus fuel balance parameter, and apply Kumar’s resulting velocity profile and propulsion control to the present mission set.)
As per claim 3, Levis does not disclose but Kumar discloses the respective first value of the balance parameter is determined in dependence on the cost for operating the respective vehicle along at least a part of the respective route, wherein said cost is dependent on one or more of fuel consumption, electrical energy consumption, battery degradation, and another degradation of the respective vehicle (.”[0072] Based on the specification data input into the exemplary embodiment of the present invention, an optimal plan which minimizes fuel use and/or emissions produced subject to speed limit constraints along the route with desired start and end times is computed to produce a trip profile 12. The profile contains the optimal speed and power (notch) settings the train is to follow, expressed as a function of distance and/or time, and such train operating limits, including but not limited to, the maximum notch power and brake settings, and speed limits as a function of location, and the expected fuel used and emissions generated.”, “[0073] The procedure used to compute the optimal profile can be any number of methods for computing a power sequence that drives the train 31 to minimize fuel and/or emissions subject to locomotive operating and schedule constraints, as summarized below. In some cases, the required optimal profile may be close enough to one previously determined, owing to the similarity of the train configuration, route and environmental conditions. In these cases, it may be sufficient to look up the driving trajectory within a database 63 and attempt to follow it. When no previously computed plan is suitable, methods to compute a new one include, but are not limited to, direct calculation of the optimal profile using differential equation models which approximate the train physics of motion. The setup involves selection of a quantitative objective function, commonly a weighted sum (integral) of model variables that correspond to rate of fuel consumption and emissions generation plus a term to penalize excessive throttle variation.”) (please see claim 1 rejection for combination rationale).
As per claim 4, Levis does not disclose establishing, in dependence on the respective first balance parameter value, for each of the vehicles a correlation set comprising a plurality of desired position and time correlations for travel of the respective vehicle along at least a part of the respective route ([0058] … The Dispatch Plan information typically includes the consignee (destination address) and associated package service levels and/or delivery commitment times ("delivery commitments"). Each group of information associated with a service stop, delivery, or other service action, can be considered a record in a database. Thus, the Dispatch Plan can be viewed as comprising a sequence of records. Further, each record could include additional information regarding customer specific requirements--e.g., certain delivery time windows )(Kumar also disclose the limitation at [0103] A requirement of the exemplary embodiment of the present invention is the ability to initially create and quickly modify on the fly any plan that is being executed. This includes creating the initial plan when a long distance is involved, owing to the complexity of the plan optimization algorithm. When a total length of a trip profile exceeds a given distance, an algorithm 46 may be used to segment the mission wherein the mission may be divided by waypoints. Though only a single algorithm 46 is discussed, those skilled in the art will readily recognize that more than one algorithm may be used where the algorithms may be connected together. The waypoint may include natural locations where the train 31 stops, such as, but not limited to, sidings where a meet with opposing traffic, or pass with a train behind the current train is scheduled to occur on single-track rail, or at yard sidings or industry where cars are to be picked up and set out, and locations of planned work. At such waypoints, the train 31 may be required to be at the location at a scheduled time and be stopped or moving with speed in a specified range. The time duration from arrival to departure at waypoints is called dwell time.).
As per claim 5, Levis discloses determining for each of the one or more of the plurality of vehicles a progress deviation indicative of a deviation of an actual progress of the respective vehicle along the respective route from a desired progress of the respective vehicle ([0090] This aspect of the Historical Data captures, in part, the "experience" aspect of a driver by way of storing past delivery information that is used to provide a benchmark to determine whether the execution of a Dispatch plan is on schedule or behind schedule. If behind schedule, there may be a need to modify (e.g., re-optimize) the remaining deliveries in the Original Dispatch Plan. For example, experienced drivers on a route benchmark their performance throughout the day by comparing their location at a known landmark with the current time, and mentally comparing these to past experience of when the landmark was encountered. Or they may compare the current time with a degree of completion of the required tasks. By comparing a delivery vehicle's current time and location relative to past average time and location measurements on that give route, a level of "experience" can be built into the system, so that a determination of the schedule status ("ahead", "behind", or "on-schedule") can be determined, as well as the time required for completion of the remaining service stops.. [0120] Another common trigger is a manual update that is entered by the user (typically the driver of the vehicle). With the manual update, the user may simply request a "check" of the status, or the user may manually add further Manifest related information. A typical embodiment is the operator requesting a status check based on the current delivery status. For example, the driver may suspect that deliveries are behind schedule and request the system to ascertain whether an updating of the Dispatch Plan is appropriate. The system then compares the current time and/or location against either the Manifest and/or historical data to obtain a benchmark as to the current delivery status.”)(Kumar also disclose the limitation at “[0178] If the operator follows the plan, the optimizer continuously displays the next operation required. If the operator does not follow the plan, the optimizer may recalculate/re-optimize the plan, depending on the deviation and the duration of the deviation of power, speed, position, emission etc. from the plan. If the operator fails to meet an optimize plan to an extent where re-optimizing the plan is not possible or where safety criteria has been or may be exceeded, in an exemplary embodiment the optimizer may take control of the vehicle to insure optimize operation, annunciate a need to consider the optimized mission plan, or simply record it for future analysis and/or use. In such an embodiment, the operator could retake control by manually disengaging the optimizer.”).
As per claim 6, Levis discloses wherein the respective progress deviation comprises a deviation, for said time correlation, of an actual position of the vehicle from a desired position according to the respective correlation set ([0101] The DM can use the current location and time to compare the location of the vehicle along a route with an expected location and time. This involves using historical data (e.g., including past delivery related times and location data) to allow the DM to determine the likelihood whether the current days' execution of the dispatch plan is on schedule, behind schedule, or ahead of schedule. In order to perform this comparison, the DM accesses a database containing historical data, including historical dispatch location and time data 36. The historical location and time data can be stored in various forms and may include a moving average of typical times associated with a given location.”). However, Levis does not disclose but Kumar discloses establishing, in dependence on the respective first balance parameter value, for each of the vehicles, a correlation set comprising a plurality of desired position and time correlations for the travel of the respective vehicle along at least a part of the respective route ([0103] A requirement of the exemplary embodiment of the present invention is the ability to initially create and quickly modify on the fly any plan that is being executed. This includes creating the initial plan when a long distance is involved, owing to the complexity of the plan optimization algorithm. When a total length of a trip profile exceeds a given distance, an algorithm 46 may be used to segment the mission wherein the mission may be divided by waypoints. Though only a single algorithm 46 is discussed, those skilled in the art will readily recognize that more than one algorithm may be used where the algorithms may be connected together. The waypoint may include natural locations where the train 31 stops, such as, but not limited to, sidings where a meet with opposing traffic, or pass with a train behind the current train is scheduled to occur on single-track rail, or at yard sidings or industry where cars are to be picked up and set out, and locations of planned work. At such waypoints, the train 31 may be required to be at the location at a scheduled time and be stopped or moving with speed in a specified range. The time duration from arrival to departure at waypoints is called dwell time.”)(please see clam 1 rejection for combination rationale).
As per claim 7, Levis does not disclose but Kumar discloses the step of obtaining for each of the one or more of the plurality of vehicles a second balance parameter value comprises determining the respective second balance parameter value in dependence on the respective progress deviation (“[0017] In another exemplary embodiment a method discloses providing an optimized mission plan to be manually applied. The optimized mission plan is re-planned in response to a manual mission plan being implemented. The manual plan is adjusted when the manual plan deviates from the optimized plan by more than a predetermined amount.”, “[0147] … This information (actual estimated arrival time or information needed to derive off-board) can also be communicated to the dispatch center to allow the dispatcher or dispatch system to adjust the target arrival times. This allows the system to quickly adjust and optimize for the appropriate target function (for example trading off speed and fuel usage). “, “[0177] FIG. 20 depicts another closed loop system where an operator is in the loop. The optimizer 650 generates the power/operating characteristic required for the optimum performance. The information is communicated to the operator 647, such as but not limited to, through human machine interface (HMI) and/or display 649. This could be in various forms including audio, text or plots or video displays. The operator 647 in this case can operate the master controller or pedals or any other actuator 651 to follow the optimum power level. [0178] If the operator follows the plan, the optimizer continuously displays the next operation required. If the operator does not follow the plan, the optimizer may recalculate/re-optimize the plan, depending on the deviation and the duration of the deviation of power, speed, position, emission etc. from the plan. If the operator fails to meet an optimize plan to an extent where re-optimizing the plan is not possible or where safety criteria has been or may be exceeded, in an exemplary embodiment the optimizer may take control of the vehicle to insure optimize operation, annunciate a need to consider the optimized mission plan, or simply record it for future analysis and/or use. In such an embodiment, the operator could retake control by manually disengaging the optimizer.”).
As per claim 10, Levis does not disclose but Kumar discloses determining, in dependence on the respective second balance parameter value, a respective velocity profile for a respective of the one or more of the plurality of vehicles for at least a portion of the respective remainder of the respective route, and controlling the respective of the one or more of the vehicles according to the respective determined velocity profile ([0072] Based on the specification data input into the exemplary embodiment of the present invention, an optimal plan which minimizes fuel use and/or emissions produced subject to speed limit constraints along the route with desired start and end times is computed to produce a trip profile 12. The profile contains the optimal speed and power (notch) settings the train is to follow, expressed as a function of distance and/or time, and such train operating limits, including but not limited to, the maximum notch power and brake settings, and speed limits as a function of location, and the expected fuel used, and emissions generated. In an exemplary embodiment, the value for the notch setting is selected to obtain throttle change decisions about once every 10 to 30 seconds. Those skilled in art will readily recognize that the throttle change decisions may occur at a longer or shorter duration, if needed and/or desired to follow an optimal speed profile.. [0074] An optimal control formulation is set up to minimize the quantitative objective function subject to constraints including but not limited to, speed limits and minimum and maximum power (throttle) settings and maximum cumulative and instantaneous emissions. Depending on planning objectives at any time, the problem may be set up flexibly to minimize fuel subject to constraints on emissions and speed limits, or to minimize emissions, subject to constraints on fuel use and arrival time. [0105] … Once travel times for each segment are allocated, a power/speed plan is determined for each segment from the previously computed solutions. If there are any waypoint constraints on speed between the segments, such as, but not limited to, a change in a speed limit, they are matched up during creation of the optimal trip profile. If speed restrictions change in only a single segment, the fuel use/travel-time curve 50 has to be re-computed for only the segment changed. This reduces time for having to re-calculate more parts, or segments, of the trip. [0106] Once a trip plan is created as discussed above, a trajectory of speed and power versus distance is used to reach a destination with minimum fuel and/or emissions at the required trip time. There are several ways in which to execute the trip plan. As provided below in more detail, in an exemplary embodiment, when in a coaching mode information is displayed to the operator for the operator to follow to achieve the required power and speed determined according to the optimal trip plan.”)
As per claim 11, Levis does not disclose but Kumar discloses obtaining a respective vehicle model in the form of a mathematical model for the respective of the one or more of the vehicles, wherein the respective velocity profile is determined by means of the respective vehicle model (“[0067] … Such input information includes, but is not limited to, train position, consist description (such as locomotive models), locomotive power description, performance of locomotive traction transmission, consumption of engine fuel as a function of output power, cooling characteristics, the intended trip route (effective track grade and curvature as function of milepost or an "effective grade" component to reflect curvature following standard railroad practices), the train represented by car makeup and loading together with effective drag coefficients, trip desired parameters including, but not limited to, start time and location, end location, desired travel time, crew (user and/or operator) identification, crew shift expiration time, and route. … [0073] …. direct calculation of the optimal profile using differential equation models which approximate the train physics of motion. The setup involves selection of a quantitative objective function, commonly a weighted sum (integral) of model variables that correspond to rate of fuel consumption and emissions generation plus a term to penalize excessive throttle variation.”) (please see claim 1 rejection for combination rationale).
As per claim 12, Levis does not disclose but Kumar discloses btaining data for the respective route, wherein the respective velocity profile is determined in dependence on the route data ([0098] A track characterization element 33 to provide information about a track, principally grade and elevation and curvature information, is also provided. The track characterization element 33 may include an on-board track integrity database 36… [0123]… As illustrated, such information is provided to an executive control element 62. Also supplied to the executive control element 62 is locomotive modeling information database 63, information from a track database 36 such as, but not limited to, track grade information and speed limit information, estimated train parameters such as, but not limited to, train weight and drag coefficients, and fuel rate tables from a fuel rate estimator 64. The executive control element 62 supplies information to the planner 12, which is disclosed in more detail in FIG. 1. Once a trip plan has been calculated, the plan is supplied to a driving advisor, driver or controller element 51. The trip plan is also supplied to the executive control element 62 so that it can compare the trip when other new data is provided.”).
As per claim 15, Levis does not disclose but Kumar discloses the respective second balance parameter value is determined in dependence on a cost for operating the vehicles at the previous set of missions (0004] This invention relates to a powered system, such as a train, an off-highway vehicle, a marine, a transport vehicle, an agriculture vehicle, and/or a stationary powered system and, more particularly to a method and computer software code for optimized fuel efficiency, emission output, vehicle performance, infrastructure and environment mission performance of the diesel powered system.. [0075] Throughout the document exemplary equations and objective functions are presented for minimizing locomotive fuel consumption. These equations and functions are for illustration only as other equations and objective functions can be employed to optimize fuel consumption or to optimize other locomotive/train operating parameters… 0083] To solve the resulting optimization problem, in an exemplary embodiment the present invention transcribes a dynamic optimal control problem in the time domain to an equivalent static mathematical programming problem with N decision variables, where the number `N` depends on the frequency at which throttle and braking adjustments are made and the duration of the trip. For typical problems, this N can be in the thousands. For example, in an exemplary embodiment, suppose a train is traveling a 172-mile (276.8 kilometers) stretch of track in the southwest United States. Utilizing the exemplary embodiment of the present invention, an exemplary 7.6% saving in fuel used may be realized when comparing a trip determined and followed using the exemplary embodiment of the present invention versus an actual driver throttle/speed history where the trip was determined by an operator. The improved savings is realized because the optimization realized by using the exemplary embodiment of the present invention produces a driving strategy with both less drag loss and little or no braking loss compared to the trip plan of the operator.)
As per claim 16, Levis does not disclose but Kumar discloses the respective second balance parameter value is determined by a control unit located remotely from the plurality of vehicles (0067] FIG. 1 depicts an exemplary illustration of a flow chart for trip optimization. As illustrated, instructions are input specific to planning a trip either on board or from a remote location, such as a dispatch center 10. Such input information includes, but is not limited to, train position, consist description (such as locomotive models), locomotive power description, performance of locomotive traction transmission, consumption of engine fuel as a function of output power, cooling characteristics, the intended trip route (effective track grade and curvature as function of milepost or an "effective grade" component to reflect curvature following standard railroad practices), the train represented by car makeup and loading together with effective drag coefficients, trip desired parameters including, but not limited to, start time and location, end location, desired travel time, crew (user and/or operator) identification, crew shift expiration time, and route.”).
Allowable Subject Matter
No prior art was applied to claims 2 and 8-9 because a combination with the prior art of record would have resulted in a piecemeal rejection using impermissible hindsight. As such, claims 2 and 8-9 are allowable.
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 OMAR ZEROUAL whose telephone number is (571)272-7255. The examiner can normally be reached Flex schedule.
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, Lynda Jasmin can be reached at (571) 272-6782. 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. Any inquiry concerning this communication or earlier communications from the examiner should be directed to OMAR ZEROUAL whose telephone number is (571)272-7255. The examiner can normally be reached Flex schedule.
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, Lynda Jasmin can be reached at (571) 272-6782. 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.
OMAR . ZEROUAL
Examiner
Art Unit 3628
/OMAR ZEROUAL/ Primary Examiner, Art Unit 3629
Prosecution Insights