Voice-Activated Smart Home Controller Using Machine Learning

The emergence of the Internet Things concept has provided a great vision for technological future, intending to enable extraction and comprehension information from environment around us, making use interaction cooperation between several devices. example Smart Homes, in particular, aims integrate these devices into households, enabling automation tasks previously performed by humans, simplify their daily lives create more comfortable environment. However, many fail keep promise, since they were not developed taking account frequent change habits tastes user, being necessary reprogramming device follow new behaviors. Taking this problem account, article presents design end-to-end implementation voice-activated smart home controller intelligent devices, deployed real validated an experimental setup motorized blinds. architecture proposed solution integrates evolvable intelligence with Online Learning framework, it automatically adapt user’s behavioral patterns. results obtained various evaluation tests provide validation operation usefulness system. main contributions work are: I) controller’s architecture; II) respective guidelines; III) open-source dataset user behavior blinds scenario; IV) comparison Offline approaches.