With increasingly complex robot hardware and the push to enable robots to work in unstructured environments, there has been an increasing interest in applying machine learning and artificial intelligence techniques to systems and control problems in robotics. At the same time, many machine learning approaches have been developed with robotic problems as motivating use cases. Robot learning spans the problems of learning models, sensing, acting, as well as integrated approaches and has been applied to many types of robot embodiments. The algorithms cover most fields of machine learning. However, due to the specific challenges in robotics, these either need to be adapted or new approaches have to be developed.
KeywordsSupervised learning Unsupervised learning Reinforcement learning End-to-end learning Model learning
- Calinon S, Lee D (2019) Chap learning control. Humanoid robotics: a reference. Springer, DordrechtGoogle Scholar
- Chatzilygeroudis K, Vassiliades V, Stulp F, Calinon S, Mouret JB (2019) A survey on policy search algorithms for learning robot controllers in a handful of trials. IEEE Trans. on Robotics – acceptedGoogle Scholar
- Osa T, Pajarinen J, Neumann G, Bagnell J, Abbeel P, Peters J (2018) An algorithmic perspective on imitation learning. Found Trends® RobotGoogle Scholar
- Premebida C, Ambrus R, Marton ZC (2018) Chap intelligent robotic perception systems. Applications of mobile robots. IntechOpen, London, pp 111–127Google Scholar
- Ruiz-del Solar J, Loncomilla P, Soto N (2018) A survey on deep learning methods for robot vision. arXiv preprint:180310862Google Scholar
- Tai L, Zhang J, Liu M, Boedecker J, Burgard W (2016) A survey of deep network solutions for learning control in robotics: from reinforcement to imitation. arXiv preprint:161207139Google Scholar