Systems and methods for dynamic pre-filtering with sampling and caching
View Patent ↗Systems and methods are provided for pre-filtering vehicle-related data obtained from vehicle sensors, V2X communications with roadside infrastructure or vehicles, and/or third-party information sources. The amount of data received from such data sources can be massive. The systems and methods pre-filter the data at the vehicle prior to transmission to an artificial intelligence or machine learning system for analysis so that the amount of data transmitted can be reduced, easing the demand on communication and data processing resources. Moreover, the speed at which the transmitted data can be analyzed is increased relative to conventional systems that rely on characterizing scenarios, training models, predicting events, etc. using as much information as can be collected.
1 . A vehicle comprising:
a sensor;
a communications circuit;
one or more processors; and
memory operatively connected to the one or more processors, and including instructions that when executed by the one or more processors, cause the vehicle to:
collect data related to a driving event from the sensors;
determine a scenario pattern for the driving event;
determine, according to the scenario pattern, a scenario pattern-specific sampling rate for sampling data points collected by the sensor for performing at least one of an extrapolation function or an interpolation function for analyzing the driving event;
extract, from the collected data, a number of data points commensurate with the scenario pattern-specific sampling rate; and
transmit, via the communications circuit, the extracted data points to an entity performing the at least one of the extrapolation function or the interpolation function for analyzing the driving event.
2 . The vehicle of claim 1 , wherein the at least one of the extrapolation function or the interpolation function generates an input to a data-reduced artificial intelligence analytics system.
3 . The vehicle of claim 1 , wherein the scenario pattern specifies the scenario pattern-specific sampling rate for performing the at least one of the extrapolation function or the interpolation function for analyzing the driving event as determined by one of a network edge device, a cloud server, or an artificial intelligence analytics system resident on the vehicle.
4 . The vehicle of claim 1 , wherein the instructions, when executed by the one or more processors, further cause the vehicle to:
receive, from the entity, an indication that a greater sampling rate is required for analyzing the driving event;
increase the scenario pattern-specific sampling rate for sampling data points collected by the sensor;
extract, from the collected data, a second number of data points commensurate with the increased scenario pattern-specific sampling rate; and
transmit the extracted second number of data points to the entity.
5 . A pre-filtering circuit of a vehicle, comprising:
one or more processors; and
memory operatively connected to the one or more processors, and including instructions that when executed by the one or more processors, cause the pre-filtering circuit to perform:
sensing first anomalous vehicle-related data;
analyzing currently collected vehicle-related data to determine existence of second anomalous vehicle-related data;
attempting to correlate the first and second anomalous vehicle-related data, and if correlatable, analyzing a series of events represented by the first and second anomalous vehicle-related data;
upon a determination that the series of events warrants further analysis, collecting additional vehicle-related data associated with the first and second anomalous vehicle-related data;
determining a scenario pattern for the series of events;
determining, according to the scenario pattern, a a scenario pattern-specific sampling rate for sampling data points collected by a sensor of the vehicle to characterize the series of events using at least one of an extrapolation function or an interpolation function; and
transmitting, to a computing entity performing the at least one of the extrapolation function or the interpolation function, the first and second anomalous vehicle-related data and an amount of the collected additional vehicle-related data collected by the sensor commensurate with the scenario pattern-specific sampling rate.
6 . The pre-filtering circuit of claim 5 , wherein the first and second anomalous vehicle-related data comprise anomalous data values relative to first and second thresholds, respectively.
7 . The pre-filtering circuit of claim 6 , wherein the first and second thresholds characterize preferred vehicle operating limits or conditions.
8 . The pre-filtering circuit of claim 5 , wherein attempting to correlate the first and second anomalous vehicle-related data comprises basing the attempted correlation on relational information pre-programmed in the pre-filtering circuit.
9 . The pre-filtering circuit of claim 5 , wherein the instructions, when executed by the one or more processors, further cause the pre-filtering circuit to perform:
receiving, from the computing entity, an indication that a greater sampling rate is required for analyzing the series of events;
increasing the scenario pattern-specific sampling rate for sampling data points collected by the sensor; and
transmitting, to the computing entity, an amount of the additional vehicle-related data collected by the sensor commensurate with the increased scenario pattern-specific sampling rate.
10 . A vehicle, comprising:
a communications circuit receiving vehicle-related data;
a pre-filtering circuit operatively connected to the communications circuit, the pre-filtering circuit:
determining whether the received vehicle-related data contains first anomalous data relative to a first vehicle operating threshold;
in response to a determination that the received vehicle-related data contains first anomalous data, triggering collection of additional vehicle-related data;
determining a scenario pattern for the first anomalous data and the additional vehicle-related data;
determining, according to the scenario pattern, a scenario pattern-specific sampling rate for sampling data points from the additional vehicle-related data to characterize one or more vehicle-related events represented by the first anomalous data and the additional vehicle-related data using at least one of an extrapolation function or an interpolation function;
extract, from the additional vehicle-related data, a number of data points commensurate with the scenario pattern-specific sampling rate;
transmitting only the first anomalous data and the extracted data points to an artificial intelligence system using the at least one of the extrapolation function or the interpolation function to characterize the one or more vehicle-related events.
11 . The vehicle of claim 10 , further comprising a plurality of operational sensors from which at least a portion of the vehicle-related data and the additional vehicle-related data originates.
12 . The vehicle of claim 10 , wherein the communications circuit receives vehicle-to-everything communications from at least one of a roadside unit and a neighboring vehicle from which at least a portion of the vehicle-related data and the additional vehicle-related data originates.
13 . The vehicle of claim 10 , wherein the pre-filtering circuit further is:
receiving, from the artificial intelligence system, an indication that a greater sampling rate is required for analyzing the one or more vehicle-related events;
increase the scenario pattern-specific sampling rate for sampling data points from the additional vehicle-related data;
extract, from the additional vehicle-related data, a second number of data points commensurate with the increased scenario pattern-specific sampling rate; and
transmit the extracted second number of data points to the artificial intelligence system.