Active Inference on the Edge: A Design Study

Abstract

Every year, the amount of data created by Internet of Things (IoT) devices increases; therefore, data processing is carried out by edge devices in close proximity. To ensure Quality of Service (QoS) throughout these operations, systems are supervised and adapted with the help of Machine Learning (ML). However, as long as ML models are not retrained, they fail to capture gradual shifts in the variable distribution, leading to an inaccurate view of the system state and poor inference. In this paper, we present a novel ML paradigm that is constructed upon Active Inference (ACI) - a concept from neuroscience that describes how the brain constantly predicts and evaluates sensory information to decrease long-term surprise. We implemented a use case, in which an ACI-based agent continuously optimized the operation on a smart manufacturing engine according to QoS requirements. The agent used causal knowledge to gradually develop an understanding of how its actions are related to requirements fulfillment, and which configurations to favor. As a result, our agent required 5 cycles to converge to the optimal solution.