DETAILED ACTION
Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim status
2. In response to the amendments filed 09/16/2025, claims 1, 2, 16 and 17 were amended and no claims were canceled and/or added. Therefore, claims 1-17 are currently pending for examination.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior office action.
Claim Rejections - 35 USC § 102
3. Claims 1-6 and 9-17 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Green et al. (Green; US 2018/0190111).
For claim 1, Green discloses a computer implemented method of controlling traffic lights according to predicted traffic patterns [E.g. 0006, 0011, 0022, 0105-0106, 0109], comprising:
receiving at least one image sequence comprising a plurality of images captured by at least one imaging sensor deployed to monitor vehicle traffic in at least one intersection [E.g. 0025: various sensors, including camera 104a-d and other sensors 106a-d, are shown to be proximate to each traffic light 102a-d. The cameras 104 capture information, such as images and other sensor data, for each arm of the intersection. In some examples, several cameras or other sensors may be placed at different locations that are not directly adjacent the traffic lights, to capture information about the intersection. In some implementations, there may be fewer sensors than one per traffic light. For example, a single sensor device or camera may capture information about several lanes, arms, or regions of a roadway so multiple sensors are not needed. In some implementations, there may be more cameras than one per traffic light, 0032: local machine learning models 112 can generate a measure of vehicles turning from a certain road segment in a certain direction. Measures of vehicles travelling in different paths can be represented as turn fractions. Turn fractions for each road segment of an intersection indicate the fraction of vehicles that have turned in a certain direction. For example, four vehicles may travel toward the intersection on the northern road segment, and three of the four vehicles may turn left onto the eastern road segment. As a result, a turn fraction for the left turns for the northern segment would be ¾ or 0.75. For each road segment connecting to an intersection, a turn fraction measure may be determined for each possible path through the intersection, e.g., a left turn fraction, a right turn fraction, a straight path fraction, etc. The total of the turn fractions for a certain road segment of an intersection is 1, which represents that the combined turn fractions represent an aggregate measure of the direction of travel of each vehicle travelling from the road segment over a period of time], at least one traffic light is deployed in the at least one intersection to control traffic flow [E.g. 0022: a traffic control system controls a traffic light at a traffic intersection. In some implementations, each traffic light is controlled by a local processing system that runs a real-time operating system (RTOS). This local processing system can be located proximate to the intersection where the controlled traffic light(s) are located. The RTOS may receive instructions from other modules in the local processing system, such as a local planning module that is part of the local processing system, and a remote planning system that sends data over a network. The local planning module may use local machine learning models to identify objects in the intersection and to generate control instructions for the traffic light based on local conditions at the intersection. The remote planning system may use remote machine learning models to generate control instructions for the traffic light based on conditions detected at many different intersections. For example, the remote planning system may receive traffic data for multiple traffic intersections and perform reinforcement learning to improve traffic flow in a large area that includes many intersections];
generating a traffic dataset descriptive of time series movement of all vehicles tracked in the at least one image sequence [E.g. 0021: In some implementations, a traffic control system includes multiple levels of machine learning processing. A remote system, e.g., a server system or cloud-based system, may use reinforcement learning to coordinate traffic control among many different traffic lights in an area such as a city. In addition, local processing systems running other machine learning models can be distributed throughout the area, for example, with a local system at each road intersection. The local systems observe conditions at their respective intersections and provide traffic data to the remote system, which uses its learning models to generate instructions that are sent to the local systems. This arrangement allows each traffic light to benefit from network-level coordination or optimization, while also retaining the flexibility for individual traffic lights and intersections to respond to local conditions with local machine learning models. For example, the instructions of the remote system may promote the overall efficiency and traffic throughput over a large area, and each local processing unit may be able to override those instructions to respond to local conditions such as an elderly person crossing the street or an emergency vehicle approaching, 0022: a traffic control system controls a traffic light at a traffic intersection. In some implementations, each traffic light is controlled by a local processing system that runs a real-time operating system (RTOS). This local processing system can be located proximate to the intersection where the controlled traffic light(s) are located. The RTOS may receive instructions from other modules in the local processing system, such as a local planning module that is part of the local processing system, and a remote planning system that sends data over a network. The local planning module may use local machine learning models to identify objects in the intersection and to generate control instructions for the traffic light based on local conditions at the intersection. The remote planning system may use remote machine learning models to generate control instructions for the traffic light based on conditions detected at many different intersections. For example, the remote planning system may receive traffic data for multiple traffic intersections and perform reinforcement learning to improve traffic flow in a large area that includes many intersections, 0053: he sensors, including cameras, may be programmed with any combination of time/day schedules, or other variables to determine whether images should be captured or not when triggers occur. The cameras may enter a low-power mode when not capturing images. In this case, the cameras may wake periodically to check for inbound messages from the controller. The cameras may be powered by internal, replaceable batteries if located remotely from the local control unit 210 or traffic light 230. The cameras may employ a small solar cell to recharge the battery when light is available. Alternatively, the cameras may be powered by the controller's 112 power supply if the cameras are co-located with the controller.];
applying a first trained machine learning model to map, based on the traffic dataset, a traffic pattern of the tracked vehicles to at least one of a plurality of learned traffic patterns [E.g. 0048: The local machine learning model 216 receives data from cameras proximate to the intersection or intersections for which the local control system 210 is responsible. In some examples, the local machine learning model 216 includes multiple machine learning models that perform different analyses. For example, the local machine learning model 216 may include a model for object identification, a model for path detection, a model for path prediction, a model for flow prediction, etc. In some examples, the local machine learning model 216 includes a single complex model that generates multiple types of traffic data.];
applying a second trained machine learning model to predict at least one future traffic flow based on the mapped traffic pattern [E.g. 0057: The remote planning system 222 includes one or more remote machine learning models 224. In some examples, the remote machine learning models 224 include one model for each type of analysis performed. For example, the remote machine learning models 224 may include a model for path pattern recognition, a model for congestion detection, a model for emergency planning and detection, etc. In some examples, the remote machine learning models 224 include one model for each subset of intersections under the purview of the remote planning system 222. For example, the remote machine learning models 224 may include one model for each state of the country, a model for each district of the state, a model for each intersection, etc.]; and
generating instructions for controlling the at least one traffic light according to the at least one predicted future traffic flow [E.g. 0022: In some implementations, a traffic control system controls a traffic light at a traffic intersection. In some implementations, each traffic light is controlled by a local processing system that runs a real-time operating system (RTOS). This local processing system can be located proximate to the intersection where the controlled traffic light(s) are located. The RTOS may receive instructions from other modules in the local processing system, such as a local planning module that is part of the local processing system, and a remote planning system that sends data over a network. The local planning module may use local machine learning models to identify objects in the intersection and to generate control instructions for the traffic light based on local conditions at the intersection. The remote planning system may use remote machine learning models to generate control instructions for the traffic light based on conditions detected at many different intersections. For example, the remote planning system may receive traffic data for multiple traffic intersections and perform reinforcement learning to improve traffic flow in a large area that includes many intersections, 0057: The remote planning system 222 includes one or more remote machine learning models 224. In some examples, the remote machine learning models 224 include one model for each type of analysis performed. For example, the remote machine learning models 224 may include a model for path pattern recognition, a model for congestion detection, a model for emergency planning and detection, etc. In some examples, the remote machine learning models 224 include one model for each subset of intersections under the purview of the remote planning system 222. For example, the remote machine learning models 224 may include one model for each state of the country, a model for each district of the state, a model for each intersection, etc.].
For claim 2, Green discloses generating the instructions for controlling the at least one traffic light according to a control plan selected based on a simulation of a plurality of control plans applied to control the at least one traffic light for controlling a flow of vehicles defined by the at least one predicted subsequent traffic flow [E.g. 0043 0048, 0061, 0070, 0076-0077, Fig. 3; system 200 uses data captured from camera system 104 at its proximate location (intersection) to manage traffic flow via traffic light manipulation, as input for simulation tools for planning ways to improve efficiency and reduce congestion].
For claim 3, Green discloses wherein the simulation is directed to predict a flow of vehicles through the at least one intersection where the selected control plan is estimated to induce optimal flow expressed by a reduced time for the vehicles to pass the at least one intersection [E.g. 0043 0048, 0061, 0070, 0076-0077, Fig. 3; system 200 uses data captured from camera system 104 at its proximate location (intersection) to manage traffic flow via traffic light manipulation, as input for simulation tools for planning ways to improve efficiency and reduce congestion].
For claim 4, Green discloses wherein the traffic dataset comprises at least one of: at least one vehicle parameter of each tracked vehicle and at least one lane parameter of each lane in the at least one intersection, the at least one vehicle parameter and the at least one lane parameter are identified based on analysis of the at least one image sequence [E.g. 0014, 0025, 0028, 0030; traffic data includes processing images captured at an intersection to determine the type of vehicle (vehicle parameter) e.g. make or model, differentiating between bus, passenger vehicles, trucks etc. as well as the number of vehicles turning (lane parameter) and travelling through an intersection, where it understood that turning will often necessitate a specific turning lane].
For claim 5, Green discloses wherein the at least one vehicle parameter is a member of a group consisting of: a vehicle type, a lane, a position in the lane, a position in a queue in the lane, a location, a relative location with respect to at least one another vehicle, a type of adjacent vehicles, a speed, an acceleration, a wait time at the at least one intersection, a distance form a stop line of the at least one intersection, and an overall tracking time [E.g. 0014, 0025, 0028, 0030; determining the type of vehicle involves determining things like the make or model or bus, passenger, truck, etc.].
For claim 6, Green discloses wherein the at least one lane parameter is a member of a group consisting of: a number of vehicles in the lane, a type of vehicles in the lane, an order of the vehicles in of a queue in the lane, a length of the queue and a lane crossing time duration [E.g. 0014, 0028; traffic data includes number of vehicles turning and travelling through an intersection].
For claim 9, Green discloses wherein the first machine learning model is trained using a plurality of traffic datasets generated based on a plurality of previously captured image sequences of the at least one intersection [E.g. 0022, 0028, 0057, 0077; local machine learning models are trained on the data from collected images regarding vehicle type, vehicle turning, crossing an intersection or approaching an intersection (plurality if sets)].
For claim 10, Green discloses wherein the first machine learning model is trained in at least one unsupervised training session to map the plurality of traffic patterns of vehicles detected at the at least one intersection to a plurality of respective clusters [E.g. 0048, 0087; local machine learning model 216 uses both clustering techniques in conjunction with an unsupervised learning approach when training to detect vehicular traffic pattern at its proximate intersection].
For claim 11, Green discloses wherein the first machine learning model is further trained post-deployment using a plurality of traffic datasets generated based on a plurality of image sequences captured after the deployment [E.g. 0087; local machine learning model 216 can be further tuned in real time (post deployment) with updated weights sent from remote planning system based on the data (images) it has collected from a plurality of local machine learning models].
For claim 12, Green discloses wherein the second machine learning model is trained in at least one supervised training session using a plurality of consecutive mapping sequences of a plurality of traffic datasets generated based on a plurality of previously captured image sequences of the at least one intersection [E.g. 0053, 0081, 0083; remote machine learning model 224 is trained on a supervised learning method with image data corresponding to different times of day or days of the week (consecutive mapping)].
For claim 13, Green discloses wherein the second machine learning model is further trained post-deployment using a plurality of consecutive mapping sequences of a plurality of traffic datasets generated based on a plurality of image sequences of the at least one intersection captured after the deployment [E.g. 0053, 0081, 0083; remote machine learning model 224 is updated (post deployment) with data from local machine learning models with data corresponding to different times of day or days of the week].
For claim 14, Green discloses wherein at least part of the process to control the at least one traffic light is executed by an edge node deployed at the at least one intersection which is functionally coupled to the at least one imaging sensor [E.g. 0021-0022; remote server system is used to coordinate traffic lights via network communication with an intersection using local processing machine (edge node) coupled to a camera system for data collection].
For claim 15, Green discloses wherein at least part of the process to control the at least one traffic light is executed by a remote server which is communicatively coupled to the at least one imaging sensor via at least one network [E.g. 0021-0022; remote server system is used to coordinate traffic light via network communication with an intersection using local processing machine (edge node) coupled to a camera system for data collection].
For claim 16, is interpreted and rejected as discussed with respect to claim 1.
For claim 17, is interpreted and rejected as discussed with respect to claim 1.
Claim Rejections - 35 USC § 103
4. Claims 7-8 are rejected under 35 U.S.C. 103 as being unpatentable over Green in view of Stein et al. (Stein; US 2019/0347821).
For claim 7, Green fails to expressly disclose wherein the analysis further comprises filtering out at least one object unrelated to tracked vehicles detected in the at least one image sequence.
However, as shown by Stein, it was well known in the art of traffic information to include an analysis that comprises filtering out at least one object unrelated to tracked vehicles detected in the at least one image sequence [0168-0169].
It would have been obvious to one of ordinary skill in the art of traffic information before
the effective filling date of the claimed invention modify Green with the teaching of Stein in order to provide the advantages of more precise counting by not being limited by clear images of vehicles by filtering out unnecessary information in captured images.
For claim 8, Green discloses wherein the analysis further comprises applying at least one trained model to track at least one vehicle [0048].
Green fails to expressly disclose applying at least one trained model to track partially visible vehicle in the at least one image sequence, the at least one partially visible vehicle is at least partially invisible in at least one of the plurality of images.
However, as shown by Stein, it was well known in the art of traffic information to include an analysis that comprises filtering out at least one object unrelated to tracked vehicles detected in the at least one image sequence [0168-0169, 0376].
It would have been obvious to one of ordinary skill in the art of traffic information before
the effective filling date of the claimed invention modify Green with the teaching of Stein in order to provide the advantages of a more robust tracking model by not being limited by clear images of vehicles by filtering out unnecessary information in captured images.
Response to Remarks
5. The Applicant's remarks regarding the rejection have been considered but they are not persuasive.
Applicant's remarks:
(1) Green fails to teach generating a traffic dataset descriptive of time series movement of all vehicles tracked in the at least one image sequence. Remarks, filed 16 September 2025, pages 8-9.
(2) Green fails to teach applying a first trained machine learning model to map a traffic pattern; applying a second trained machine learning model to predict at least one future traffic flow based on the mapped traffic pattern. Remarks, filed 16 September 2025, pages 8-9.
(3) Green is a Non Enabling Art. Remarks, filed 16 September 2025, page 10.
(4) As to claims 7-8, Green and Stein fail to teach the limitations of claim 4 on which it depends that requires. Remarks, filed 16 September 2025, page 11.
(5) As to claim 8 Green and Stein fail to teach applying at least one trained model to track at least one partially visible vehicle in the at least one image sequence. Remarks, filed 16 September 2025, page 11.
(6) The combination used in rejection claim 7 is improper and the rejections be removed. Remarks, filed 16 September 2025, page 12.
(7) Applicant respectfully submits that the combination of Green and Stein is improper because the Office Action relies on information gleaned solely from Applicant’s specification. Remarks, filed 16 September 2025, pages 12-13.
Examiner’s response:
Regarding Applicant first remark, Green expressly disclose data is gathered (generated) related to traffic condition such as volume, flow, lane, etc. (descriptive) during various time of day [E.g. 0021-0023, 0053]. One of ordinary skill in the art would consider that generating traffic dataset of all vehicles tracked in the at least one image sequence using sensors during various times of the day is considered generating the traffic dataset as a time series movement.
Regarding Applicant second remark, Green expressly disclose a applying a first trained machine learning model to map a traffic pattern [a local (first trained) machine learning model 216 makes (maps) traffic flow predication (pattern) using data gathered on traffic movement at various times of day at its proximate intersection; 0048, 0053]; applying a second trained machine learning model to predict at least one future traffic flow based on the mapped traffic pattern [remote (second trained) machine learning models 224 use data from local machine learning model for traffic pattern recognition; the data is transmitted from local machine learning model 216 to remote machine learning model 224; 0057, 0055-0056].
Regarding Applicant third remark, it has been held that a prior art reference must either be in the field of the inventor' s endeavor or, if not, then be reasonably pertinent to the particular problem with which the inventor was concerned, in order to be relied upon as a basis for rejection of the claimed invention. See In re Oetiker, 977 F.2d 1443, 24 USPQ2d 1443 (Fed. Cir. 1992).
Regarding Applicant fourth remark, as discussed in and analysis of the claims Green discloses wherein the traffic dataset comprises at least one of: at least one vehicle parameter of each tracked vehicle and at least one lane parameter of each lane in the at least one intersection, the at least one vehicle parameter and the at least one lane parameter are identified based on analysis of the at least one image sequence [E.g. 0014, 0025, 0028, 0030; traffic data includes processing images captured at an intersection to determine the type of vehicle (vehicle parameter) e.g. make or model, differentiating between bus, passenger vehicles, trucks etc. as well as the number of vehicles turning (lane parameter) and travelling through an intersection, where it understood that turning will often necessitate a specific turning lane].
Furthermore, Applicant's arguments fail to comply with 37 CFR 1.111(b) because they amount to a general allegation that the claims define a patentable invention without specifically pointing out how the language of the claims patentably distinguishes them from the references.
Regarding Applicant fifth remark Green discloses wherein the analysis further comprises applying at least one trained model to track at least one vehicle [data gathered on traffic movement at various times of day at its proximate intersections; 0048].
Green just fails to expressly disclose applying at least one trained model to track partially visible vehicle in the at least one image sequence, the at least one partially visible vehicle is at least partially invisible in at least one of the plurality of images.
Stein is relied on to show that it was well known in the art of traffic information to include applying at least one trained model to track partially visible vehicle in the at least one image sequence, the at least one partially visible vehicle is at least partially invisible in at least one of the plurality of images [target (tracked) vehicles via multiple frames of images, may be partially obstructed (invisible) from imaging sensor but the location may be inferred using from estimated relevant distance from known vehicle or path trajectory estimation; 0168-0169, 0376].
It would have been obvious to one of ordinary skill in the art of traffic information before
the effective filling date of the claimed invention modify Green with the teaching of Stein in order to provide the advantages of a more robust tracking model by not being limited by clear images of vehicles by filtering out unnecessary information in captured images.
Regarding Applicant sixth remark, the examiner recognizes that obviousness may be established by combining or modifying the teachings of the prior art to produce the claimed invention where there is some teaching, suggestion, or motivation to do so found either in the references themselves or in the knowledge generally available to one of ordinary skill in the art. See In re Fine, 837 F.2d 1071, 5 USPQ2d 1596 (Fed. Cir. 1988), In re Jones, 958 F.2d 347, 21 USPQ2d 1941 (Fed. Cir. 1992), and KSR International Co. v. Teleflex, Inc., 550 U.S. 398, 82 USPQ2d 1385 (2007).
Regarding Applicant seventh remark, it must be recognized that any judgment on obviousness is in a sense necessarily a reconstruction based upon hindsight reasoning. But so long as it takes into account only knowledge which was within the level of ordinary skill at the time the claimed invention was made, and does not include knowledge gleaned only from the applicant's disclosure, such a reconstruction is proper. See In re McLaughlin, 443 F.2d 1392, 170 USPQ 209 (CCPA 1971).
Conclusion
6. THIS ACTION IS MADE FINAL. 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.
7. Any inquiry concerning this communication or earlier communications from the examiner should be directed to MOHAMED BARAKAT whose telephone number is (571)270-3696. The examiner can normally be reached on 9:00am-5:00PM.
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, Davetta Goins can be reached on (571) 272-2957. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MOHAMED BARAKAT/
Primary Examiner, Art Unit 2689