Big Data Analytics as a Service for Affective Humanoid Service Robots

摘要

This paper identifies and analyses the advanced capability requirements for humanoid service robots to serve in highly complicated and intelligence demanding applications, such as children education and home care, in future smart home environments. In particular, a Distributed Collaboration and Continuous Learning (DCCL) mechanism is identified as the key capability of a humanoid service robot to succeed in these applications. Based on the latest Big Data Analytics tools with distributed machine learning technologies integrated as services, a novel DCCL middleware platform is developed to facilitate the realisation of the DCCL mechanism. A user preference based children toy recommendation application is introduced as a use case study of the DCCL mechanism and platform.