Session: MR-05-01 - Motion Planning, Dynamics, and Control of Robots

Paper Number: 88459

88459 - Metareasoning Approaches for Thermal Management During Image Processing 

Resource-constrained electronic systems are present in many semi- and fully-autonomous systems and are tasked with computationally heavy tasks such as neural network image processing. Without sufficient cooling, these tasks often increase device temperature up to a predetermined maximum, beyond which the task is slowed by the device firmware to maintain the maximum. This is done to avoid decreased processor lifespan due to thermal fatigue or catastrophic processor failure due to thermal overstress. This paper describes a study that evaluated how well metareasoning can manage a Raspberry Pi 4B’s central processing unit (CPU) temperature while it is performing image processing (object detection and classification) on the Common Objects in Context (COCO) dataset. Two methods which employ metareasoning were developed; one which maintains constant image throughput, and another which maintains constant expected detection accuracy. Method one utilized both the InceptionV2 and MobileNetV2 image classification networks with a Single Shot Multibox Detector (SSD) attached. These networks were alternated depending on the current CPU temperature. Method two was tested on each network for a range of parameter values. The study also considered cases that used the system's built-in throttling method to control the temperature. Both metareasoning approaches were able to stabilize the device temperature without relying on throttling. 

Presenting Author: Michael Dawson University of Maryland, College Park

Presenting Author Biography: Michael Dawson is a student at the University of Maryland, College Park. He is pursing his M.S. in Mechanical Engineering with an expected graduation of May, 2022. His interests lie in metareasoning, control, automation, and autonomous systems.

Authors:

Michael Dawson University of Maryland, College Park
Jeffrey Herrmann University of Maryland, College Park