State estimation and response to active school vehicles in a self-driving system
This document discloses system, method, and computer program product embodiments for anticipating an imminent state of a school transportation vehicle. For example, the method includes receiving sensor data of an environment near an autonomous vehicle. The method further includes, in response to the sensor data including a representation of a school transportation vehicle, analyzing the sensor data to estimate, from a set of candidate states, a current state of the school transportation vehicle, wherein the candidate states include an actively loading or unloading state, an imminently loading or unloading state, and an inactive state. The method further includes, in response to the estimated current state being either the actively loading or unloading state or the imminently loading or unloading state, causing the autonomous vehicle to slow or stop until the school transportation vehicle is in the inactive state.
1 . A method of anticipating an imminent state of a school transportation vehicle comprising:
receiving sensor data of an environment near an autonomous vehicle;
in response to the sensor data including a representation of a school transportation vehicle:
analyzing the sensor data to estimate, based on indicators from the school transportation vehicle and indicators from a surrounding environment of the school transportation vehicle, from a set of candidate states, a current state of the school transportation vehicle, wherein the candidate states comprise an actively loading or unloading state, an imminently loading or unloading state, and an inactive state; and
in response to the estimated current state being either the actively loading or unloading state or the imminently loading or unloading state, causing the autonomous vehicle to slow or stop until the school transportation vehicle is in the inactive state.
2 . The method of claim 1 , wherein causing the autonomous vehicle to slow or stop comprises imposing a goal on a motion control system of the autonomous vehicle based on the estimated state of the school transportation vehicle, and the indicators from a surrounding environment of the school transportation vehicle includes at least one of: pedestrian presence, and other stopping vehicles.
3 . The method of claim 2 , wherein the imposed goal is based on a traffic regulation associated with the school transportation vehicle.
4 . The method of claim 1 , wherein:
each candidate state has an associated set of indicators; and
analyzing the sensor data comprises detecting one or more of the associated indicators in the sensor data.
5 . The method of claim 4 , wherein analyzing the sensor data further comprises computing a probability mass function for the candidate states based on the detected indicators.
6 . The method of claim 5 , wherein causing the autonomous vehicle to slow or stop comprises imposing a goal on a motion control system of the autonomous vehicle based on the probability mass function for the candidate states.
7 . The method of claim 1 , wherein analyzing the sensor data to estimate the current state of the school transportation vehicle further comprises analyzing a calendar date and/or a time of day to estimate if a school transportation vehicle is in the actively loading or unloading state, the imminently loading or unloading state, or the inactive state.
8 . The method of claim 1 , further comprising:
detecting, based on a subsequent image, a subsequent state of the school transportation vehicle; and
determining whether the subsequent state corresponds to the anticipated imminent state.
9 . A system, comprising:
a memory; and
at least one processor coupled to the memory and configured to:
receive sensor data of an environment near an autonomous vehicle;
in response to the sensor data including a representation of a school transportation vehicle:
analyze the sensor data to estimate based on indicators from the school transportation vehicle and indicators from a surrounding environment of the school transportation vehicle, from a set of candidate states, a current state of the school transportation vehicle, wherein the candidate states comprise an actively loading or unloading state, an imminently loading or unloading state, and an inactive state; and
in response to the estimated current state being either the actively loading or unloading state or the imminently loading or unloading state, cause the autonomous vehicle to slow or stop until the school transportation vehicle is in the inactive state.
10 . The system of claim 9 , wherein the at least one processor is configured to cause the autonomous vehicle to slow or stop by imposing a goal on a motion control system of the autonomous vehicle based on the estimated state of the school transportation vehicle, and the indicators from the school transportation vehicle includes a speed of the school transportation vehicle.
11 . The system of claim 10 , wherein the imposed goal is based on a traffic regulation associated with the school transportation vehicle.
12 . The system of claim 9 , wherein:
each candidate state has an associated set of indicators; and
the at least one processor is configured to analyze the sensor data by:
detecting one or more of the associated indicators in the sensor data and analyzing the sensor data; and
computing a probability mass function for the candidate states based on the detected indicators.
13 . The system of claim 12 , wherein the at least one processor is configured to cause the autonomous vehicle to slow or stop by imposing a goal on a motion control system of the autonomous vehicle based on the probability mass function for the candidate states.
14 . A non-transitory computer-readable medium that stores instructions that are configured to, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:
receiving sensor data of an environment near an autonomous vehicle;
in response to the sensor data including a representation of a school transportation vehicle:
analyzing the sensor data to estimate based on indicators from the school transportation vehicle and indicators from a surrounding environment of the school transportation vehicle, from a set of candidate states, a current state of the school transportation vehicle, wherein the candidate states comprise an actively loading or unloading state, an imminently loading or unloading state, and an inactive state; and
in response to the estimated current state being either the actively loading or unloading state or the imminently loading or unloading state, causing the autonomous vehicle to slow or stop until the school transportation vehicle is in the inactive state.
15 . The non-transitory computer-readable medium of claim 14 , wherein causing the autonomous vehicle to slow or stop comprises imposing a goal on a motion control system of the autonomous vehicle based on the estimated state of the school transportation vehicle, and the indicators from a surrounding environment of the school transportation vehicle includes a presence of passengers inside the school transportation vehicle.
16 . The non-transitory computer-readable medium of claim 15 , wherein the imposed goal is based on a traffic regulation associated with the school transportation vehicle.
17 . The non-transitory computer-readable medium of claim 14 , wherein:
each candidate state has an associated set of indicators; and
analyzing the sensor data comprises detecting one or more of the associated indicators in the sensor data.
18 . The non-transitory computer-readable medium of claim 17 , wherein analyzing the sensor data further comprises computing a probability mass function for the candidate states based on the detected indicators.
19 . The non-transitory computer-readable medium of claim 18 , wherein causing the autonomous vehicle to slow or stop comprises imposing a goal on a motion control system of the autonomous vehicle based on the probability mass function for the candidate states.
20 . The non-transitory computer-readable medium of claim 14 , wherein analyzing the sensor data to estimate the current state of the school transportation vehicle further comprises analyzing a calendar date and/or a time of day.