Intelligent railroad at-grade crossings
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for monitoring vehicles traversing a dedicated roadway that includes an at-grade crossing. In some implementations, a system includes a central server, a gate system, and sensors. The gate system provides access to an at-grade crossing for vehicles. The sensors are positioned in a fixed location relative to a roadway, the roadway including the at-grade crossing. Each sensor can detect vehicles on the roadway. For each vehicle, each sensor can generate sensor data and observational data from the generated sensor data. Each sensor can determine a likelihood that the detected vehicle will approach the at-grade crossing by comparing the likelihood to a threshold. In response, each sensor can transmit data to the gate system that causes the gate system to allow the autonomous vehicle access to the at-grade crossing prior to the autonomous vehicle reaching the gate system.
1. A computer-implemented method comprising:
receiving, by a first sensor device that is positioned to observe vehicles at a first location along a roadway, a received feature set from a second sensor device that is positioned to observe vehicles at a different location along the roadway, the received feature set indicating one or more characteristics of a vehicle as sensed by the second sensor device;
generating, by the first sensor device, a generated feature set, the generated feature set indicating one or more characteristics of the vehicle as sensed by the first sensor device;
determining, by the first sensor device, that the received feature set matches the generated feature set;
after determining that the received feature set matches the generated feature set, by the first sensor device, that a model indicates that vehicle, determining, by the first sensor device, that the vehicle is likely to approach an intersection; and
in response to determining that the vehicle is likely to approach the intersection, transmitting, by the first sensor device, a message to an access control device that controls access to the intersection.
2. The method of claim 1 , wherein the access control device comprises a gate.
3. The method of claim 1 , wherein the intersection comprises an intersection of two roadways.
4. The method of claim 1 , wherein the determining that the vehicle is likely to approach the intersection comprises inputting the received feature set or the generated feature set to a machine learning model.
5. The method of claim 1 , comprising:
determining whether the vehicle actually approached the intersection; and
updating a machine learning model based on determining whether the vehicle actually approached the intersection.
6. The method of claim 1 , wherein the received feature set indicates one or more characteristics of the vehicle as sensed by a LIDAR unit or a radar unit of the second sensor device.
7. The method of claim 1 , wherein the received feature set indicates one or more characteristics of the vehicle as sensed by a microphone of the second sensor device.
8. The method of claim 1 , wherein the received feature set indicates one or more characteristics of the vehicle as sensed by a camera of the second sensor device.
9. The method of claim 1 , wherein determining that the received feature set matches the generated feature set comprises comparing an identity product of the received feature set with an identity product of the generated feature set.
10. The method of claim 1 , comprising, after determining that the received feature set matches the generated feature set, transmitting, by the first sensor device, a message to the second sensor device.
11. The method of claim 1 , wherein a field of view of the first sensor device at the first location along the roadway is non-overlapping with a field of view of the second sensor device at the different location along the roadway.
12. One or more non-transitory computer-readable media that store instructions which, when executed by one or more computer processors, cause the one or more computer processors to perform operations comprising:
receiving, by a first sensor device that is positioned to observe vehicles at a first location along a roadway, a received feature set from a second sensor device that is positioned to observe vehicles at a different location along the roadway, the received feature set indicating one or more characteristics of a vehicle as sensed by the second sensor device;
generating, by the first sensor device, a generated feature set, the generated feature set indicating one or more characteristics of the vehicle as sensed by the first sensor device;
determining, by the first sensor device, that the received feature set matches the generated feature set;
after determining that the received feature set matches the generated feature set, by the first sensor device, that a model indicates that vehicle, determining, by the first sensor device, that the vehicle is likely to approach an intersection; and
in response to determining that the vehicle is likely to approach the intersection, transmitting, by the first sensor device, a message to an access control device that controls access to the intersection.
13. The media of claim 12 , wherein the access control device comprises a gate.
14. The media of claim 12 , wherein the intersection comprises an intersection of two roadways.
15. The media of claim 12 , wherein the determining that the vehicle is likely to approach the intersection comprises inputting the received feature set or the generated feature set to a machine learning model.
16. The media of claim 12 , wherein the operations comprise:
determining whether the vehicle actually approached the intersection; and
updating a machine learning model based on determining whether the vehicle actually approached the intersection.
17. The media of claim 12 , wherein the received feature set indicates one or more characteristics of the vehicle as sensed by a LIDAR unit or a radar unit of the second sensor device.
18. The media of claim 12 , wherein the received feature set indicates one or more characteristics of the vehicle as sensed by a microphone of the second sensor device.
19. The media of claim 12 , wherein the received feature set indicates one or more characteristics of the vehicle as sensed by a camera of the second sensor device.
20. A system comprising:
one or more computer processors; and
one or more non-transitory computer-readable media that store instructions which, when executed by the one or more computer processors, cause the one or more computer processors to perform operations comprising:
receiving, by a first sensor device that is positioned to observe vehicles at a first location along a roadway, a received feature set from a second sensor device that is positioned to observe vehicles at a different location along the roadway, the received feature set indicating one or more characteristics of a vehicle as sensed by the second sensor device;
generating, by the first sensor device, a generated feature set, the generated feature set indicating one or more characteristics of the vehicle as sensed by the first sensor device;
determining, by the first sensor device, that the received feature set matches the generated feature set;
after determining that the received feature set matches the generated feature set, by the first sensor device, that a model indicates that vehicle, determining, by the first sensor device, that the vehicle is likely to approach an intersection; and
in response to determining that the vehicle is likely to approach the intersection, transmitting, by the first sensor device, a message to an access control device that controls access to the intersection.