Indexing sensor data about the physical world
Systems, methods, and other embodiments described herein relate to improving the use of sensor data in a mobile context by indexing the sensor data with derived content. In one embodiment, a method includes, responsive to an offload event, acquiring sensor data from a vehicle. The method includes generating an index of the sensor data according to attributes of how the sensor data was acquired. The method includes analyzing the sensor data to derive additional content using at least a model. The method includes updating the index using the additional content to further indicate characteristics of the sensor data relative to the attributes. The method includes providing the index to a remote device as a report about contents of the sensor data.
1 . A collection system for improving offloading and storage of sensor data, comprising:
one or more processors; and
a memory communicably coupled to the one or more processors and storing instructions that, when executed by the one or more processors, cause the one or more processors to:
responsive to an offload event, acquire sensor data from a vehicle at an edge device;
generate, at the edge device, an index of the sensor data according to attributes of how the sensor data was acquired, including derived perceptions from automated driving systems and safety control systems of the vehicle;
analyze, at the edge device, the sensor data to derive additional content using at least a model;
update, at the edge device, the index using the additional content to further indicate characteristics of the sensor data relative to the attributes; and
provide the index to a remote device as a report about contents of the sensor data.
2 . The collection system of claim 1 , wherein the instructions to analyze the sensor data include instructions to apply, by the edge device, the model to derive perceptions about a surrounding environment of the vehicle represented by the sensor data including a presence of objects, locations of the objects, types of the objects, and relationships between the objects.
3 . The collection system of claim 1 , wherein the instructions to generate the index include instructions to correlate the sensor data with the attributes about acquisition of the sensor data including at least a time and a location for when respective segments of the sensor data were acquired.
4 . The collection system of claim 3 , wherein the instructions to update the index include instructions to correlate the additional content with the sensor data in the index according to the time and the location of the respective segments associated with the additional content, and
wherein the characteristics include content of the sensor data that defines scenarios associated with a combination of objects, locations of the objects, and a roadway context.
5 . The collection system of claim 1 , wherein the instructions to provide the index include instructions to communicate the index to a cloud-based device to aggregate the index with indexes from other vehicles in order to form a combined catalog of available sensor data among the vehicle and the other vehicles over a broad geographic area.
6 . The collection system of claim 1 , wherein the instructions to generate the index according to the attributes include instructions to index the sensor data by at least a geographic location of where the sensor data was acquired and a time of when the sensor data was acquired, and
wherein the instructions to generate the index include instructions to populate the index with the derived perceptions produced by the vehicle associated with safety systems of the vehicle, and automated control systems of the vehicle.
7 . The collection system of claim 1 , wherein the instructions to acquire the sensor data further include instructions to acquire the derived perceptions about the sensor data from the vehicle and sampling the sensor data to compress the sensor data for storage.
8 . The collection system of claim 1 , wherein the offload event is identified and performed from a group including detection of the vehicle within a communication proximity of the edge device and charging of the vehicle at the edge device.
9 . A non-transitory computer-readable medium storing instructions for improving offloading and storage of sensor data and that, when executed by one or more processors, cause the one or more processors to:
responsive to an offload event, acquire sensor data from a vehicle, at an edge device;
generate, at the edge device, an index of the sensor data according to attributes of how the sensor data was acquired;
analyze, at the edge device, the sensor data to derive additional content using at least a model;
update, at the edge device, the index using the additional content to further indicate characteristics of the sensor data relative to the attributes; and
provide the index to a remote device as a report about contents of the sensor data.
10 . The non-transitory computer-readable medium of claim 9 , wherein the instructions to analyze the sensor data include instructions to apply, by the edge device, the model to derive perceptions about a surrounding environment of the vehicle represented by the sensor data including a presence of objects, locations of the objects, types of the objects, and relationships between the objects.
11 . The non-transitory computer-readable medium of claim 9 , wherein the instructions to generate the index include instructions to correlate the sensor data with the attributes about acquisition of the sensor data including at least a time and a location for when respective segments of the sensor data were acquired.
12 . The non-transitory computer-readable medium of claim 11 , wherein the instructions to update the index include instructions to correlate the additional content with the sensor data in the index according to the time and the location of the respective segments associated with the additional content, and
wherein the characteristics include content of the sensor data that defines scenarios associated with a combination of objects, locations of the objects, and a roadway context.
13 . The non-transitory computer-readable medium of claim 9 , wherein the instructions to provide the index include instructions to communicate the index to a cloud-based device to aggregate the index with indexes from other vehicles in order to form a combined catalog of available sensor data among the vehicle and the other vehicles over a broad geographic area.
14 . A method, comprising:
responsive to an offload event, acquiring sensor data from a vehicle at an edge device;
generating, at the edge device, an index of the sensor data according to attributes of how the sensor data was acquired, including derived perceptions from automated driving systems and safety control systems of the vehicle;
analyzing, at the edge device, the sensor data to derive additional content using at least a model;
updating, at the edge device, the index using the additional content to further indicate characteristics of the sensor data relative to the attributes; and
providing the index to a remote device as a report about contents of the sensor data.
15 . The method of claim 14 , wherein analyzing the sensor data includes applying the model, by the edge device, to derive perceptions about a surrounding environment of the vehicle represented by the sensor data including a presence of objects, locations of the objects, types of the objects, and relationships between the objects.
16 . The method of claim 14 , wherein generating the index includes correlating the sensor data with the attributes about acquisition of the sensor data including at least a time and a location for when respective segments of the sensor data were acquired.
17 . The method of claim 16 , wherein updating the index includes correlating the additional content with the sensor data in the index according to the time and the location of the respective segments associated with the additional content, and
wherein the characteristics include content of the sensor data that defines scenarios associated with a combination of objects, locations of the objects, and a roadway context.
18 . The method of claim 14 , wherein providing the index includes communicating the index to a cloud-based device to aggregate the index with indexes from edge devices about other vehicles in order to form a combined catalog of available sensor data among the edge devices over a broad geographic area.
19 . The method of claim 14 , wherein generating the index according to the attributes includes indexing the sensor data by at least a geographic location of where the sensor data was acquired and a time of when the sensor data was acquired, and
wherein generating the index includes populating the index with the derived perceptions produced by the vehicle associated with safety systems of the vehicle, and automated control systems of the vehicle.
20 . The method of claim 14 , wherein acquiring the sensor data further includes acquiring derived information about the sensor data from the vehicle and sampling the sensor data to compress the sensor data for storage, and
wherein the offload event is identified and performed from a group including detection of the vehicle within a communication proximity of the edge device and charging of the vehicle at the edge device.