When and how to help: An iterative probabilistic model for learning assistance by demonstration

Abstract

Abstract-Crafting a proper assistance policy is a difficult endeavour but essential for the development of robotic assistants. Indeed, assistance is a complex issue that depends not only on the task-at-hand, but also on the state of the user, environment and competing objectives. As a way forward, this paper proposes learning the task of assistance through observation; an approach we term Learning Assistance by Demonstration (LAD). Our methodology is a subclass of Learningby-Demonstration (LbD), yet directly addresses difficult issues associated with proper assistance such as when and how to appropriately assist. To learn assistive policies, we develop a probabilistic model that explicitly captures these elements and provide efficient, online, training methods. Experimental results on smart mobility assistance -using both simulation and a real-world smart wheelchair platform -demonstrate the effectiveness of our approach; the LAD model quickly learns when to assist (achieving an AUC score of 0.95 after only one demonstration) and improves with additional examples. Results show that this translates into better task-performance; our LAD-enabled smart wheelchair improved participant driving performance (measured in lap seconds) by 20.6s (a speedup of 137%), after a single teacher demonstration.

The Transfer of Human Trust in Robot Capabilities across Tasks

“…As such, we develop models that capture both the dynamic property of trust and its variation across tasks. We leverage upon recurrent neural networks that have been applied to a variety of sequential learning tasks (e.g., [34]) and online Gaussian processes that have been previously used in robotics [33,31,32].…”

Online Multimodal Ensemble Learning Using Self-Learned Sensorimotor Representations

“…One possibility to acquire a target trajectory and to learn inverse models from data is to collect multiple demonstrations of each task. This approach has been explored for example in [39][40][41][42][43], where statistical methods have been proposed to infer models from multiple demonstrated trajectory. However, this approach requires multiple demonstrations, which might be difficult to obtain.…”

One-shot assistance estimation from expert demonstrations for a shared control wheelchair system

“…The fundamental point underlying such methodologies is that assistive systems should not unconditionally assist the user at all times, but attempt to infer the current needs of the user, and act accordingly [9]. We suggest that an assistive robot would be most effective if it can emulate the operation of a human assistant, e.g.…”

“…the assistant) and the user, while keeping track of what is going on within the environment. Up to our knowledge, such complex scenarios, where the demonstrations are given by a human assistant while the user and the robot pursues interaction, has not yet been considered except for a recent work by Soh and Demiris [9]. This paper is organized as follows: Section II presents related work on intelligent wheelchair systems.…”

“…That is why almost all existing wheelchair systems implement restrictive guidance schemes for strict trajectory-following tasks. A recent study by Soh and Demiris proposed that a robotic wheelchair can learn from the demonstrations of another person, who can present the user human guidance during navigation by taking over the control of the wheelchair [9], [25]. Assuming the coherence of human guidance, the robotic wheelchair modeled the user's actions and learned how and when to assist the user.…”