Arcades: A deep model for adaptive decision making in voice controlled smart-home

TL;DR

Arcades is a deep reinforcement learning system that enables adaptive, context-aware decision making in voice-controlled smart-homes, handling environmental changes and improving long-term robustness.

Contribution

The paper introduces Arcades, a novel deep reinforcement learning approach that uses graphical representations for robust, adaptive smart-home control.

Findings

Demonstrates robustness to sensor failures and environmental changes.

Shows promising results for long-term context-aware smart-home management.

Utilizes deep RL to adapt continuously to user interactions.

Abstract

In a voice-controlled smart-home, a controller must respond not only to user's requests but also according to the interaction context. This paper describes Arcades, a system which uses deep reinforcement learning to extract context from a graphical representation of home automation system and to update continuously its behavior to the user's one. This system is robust to changes in the environment (sensor breakdown or addition) through its graphical representation (scale well) and the reinforcement mechanism (adapt well). The experiments on realistic data demonstrate that this method promises to reach long life context-aware control of smart-home.