Controlling Concurrent Execution of Perception Algorithms

A mobile computing device for controlling concurrent execution of perception algorithms, the mobile computing device comprising:

• network interface circuitry, and processing circuitry,
• the mobile computing device operative to:
• acquire sensor data from one or more sensors, the acquired sensor data representing a physical environment in which the mobile computing device is located,
• execute an application which is dependent on one or more perception algorithms processing the acquired sensor data, the one or more perception algorithms being subject to respective perception-performance constraints and a power-consumption constraint,
• determine operating conditions of the mobile computing device which have an impact on perception performance and power consumption of the one or more perception algorithms,
• estimate a level of contention for a local computing resource on which at least one of the one or more perception algorithms and the application, or at least two of the perception algorithms, can be executed concurrently,
• estimate a probability of failure for at least one of the one or more perception algorithms based on the estimated level of contention, respective current configurations of the one or more perception algorithms, a current configuration of the local computing resource, and the determined operating conditions,
• if the estimated probability of failure exceeds a failure threshold, offload execution of the at least one perception algorithm to an external computing resource, and
• if the estimated probability of failure does not exceed the failure threshold, execute the at least one perception algorithm on the local computing resource.

A method of controlling concurrent execution of perception algorithms, the method performed by a mobile computing device and comprising:

• acquiring sensor data from one or more sensors, the acquired sensor data representing a physical environment in which the mobile computing device is located,
• executing an application which is dependent on one or more perception algorithms processing the acquired sensor data, the one or more perception algorithms being subject to respective perception-performance constraints and a power-consumption constraint,
• determining operating conditions of the mobile computing device which have an impact on perception performance and power consumption of the one or more perception algorithms,
• estimating a level of contention for a local computing resource comprised in the mobile computing device, on which local computing resource at least one of the one or more perception algorithms and the application, or at least two of the perception algorithms, can be executed concurrently,
• estimating a probability of failure for at least one of the one or more perception algorithms based on the estimated level of contention, respective current configurations of the one or more perception algorithms, a current configuration of the local computing resource, and the determined operating conditions, and
• offloading execution of the at least one perception algorithm to an external computing resource, or executing the at least one perception algorithm on the local computing resource, depending respectively on whether or not the estimated probability of failure exceeds a failure threshold.

A non-transitory computer-readable medium on which is stored a computer program comprising instructions which, when the computer program is executed by a mobile computing device, causes the mobile computing device to:

• acquire sensor data from one or more sensors, the acquired sensor data representing a physical environment in which the mobile computing device is located,
• execute an application which is dependent on one or more perception algorithms processing the acquired sensor data, the one or more perception algorithms being subject to respective perception-performance constraints and a power-consumption constraint,
• determine operating conditions of the mobile computing device which have an impact on perception performance and power consumption of the one or more perception algorithms,
• estimate a level of contention for a local computing resource on which at least one of the one or more perception algorithms and the application, or at least two of the perception algorithms, can be executed concurrently,
• estimate a probability of failure for at least one of the one or more perception algorithms based on the estimated level of contention, respective current configurations of the one or more perception algorithms, a current configuration of the local computing resource, and the determined operating conditions,
• if the estimated probability of failure exceeds a failure threshold, offload execution of the at least one perception algorithm to an external computing resource, and
• if the estimated probability of failure does not exceed the failure threshold, execute the at least one perception algorithm on the local computing resource.