User-adapted plan recognition and user-adapted shared control: A Bayesian approach to semi-autonomous wheelchair driving

Abstract

Many elderly and physically impaired people experience difficulties when maneuvering a powered wheelchair. In order to ease maneuvering, powered wheelchairs have been equipped with sensors, additional computing power and intelligence by various research groups.This paper presents a Bayesian approach to maneuvering assistance for wheelchair driving, which can be adapted to a specific user. The proposed framework is able to model and estimate even complex user intents, i.e. wheelchair maneuvers that the driver has in mind. Furthermore, it explicitly takes the uncertainty on the user's intent into account. Besides during intent estimation, user-specific properties and uncertainty on the user's intent are incorporated when taking assistive actions, such that assistance is tailored to the user's driving skills. This decision making is modeled as a greedy Partially Observable Markov Decision Process (POMDP).Benefits of this approach are shown using experimental results in simulation and on our wheelchair platform Sharioto.

Learning user habits for semi-autonomous navigation using low throughput interfaces

“…Shared-control systems normally require constant user involvement in order to provide navigational commands [8], [14]; while standard semi-autonomous systems also rely on several input signals for selecting a command through some user interface (such as going up or down in a predefined menu or selecting an item, [15]) or for providing directions to the system (left, right, forward) at a given point in the travel [16].…”

“…going left/right/forward, passing through a door, or avoiding an obstacle. Among these, dynamic Bayesian networks (DBNs) have been used to distinguish between approaching or avoiding an obstacle, or to control a wheelchair using either a joystick or brain-decoded commands [2], [8]. Alternatively, Carlson and Demiris [9] computed the probability of trajectories obtained from the relative pose and orientation between a wheelchair and two predefined possible goals.…”

Breaking the Human-Robot Deadlock: Surpassing Shared Control Performance Limits with Sparse Human-Robot Interaction

“…For example, in [28], the optimal controller of a cart-pendulum task was used in conjunction with a haptic interface to filter suboptimal user input (the machine decided when to give assistance). Some researchers, such as [17], [34], [18], infer user intent as a maximum likelihood, maximum a-posteriori, or partially observable Markov decision process estimation problem; in turn, shared control is computed as a reward function over the optimal user action and optimal robot action Figure 1 could be a reward function).…”

Adaptive Aid on Targeted Robot Manipulator Movements in Tele-Assistance

“…This is a similar concept to "shared control", where the user and the robot use their own sensing, control and planning capabilities in a cooperative way. Examples of assistive technologies include wheelchairs that attempt to aid the disabled or elderly user in performing navigation tasks [17,18]. Shared control has also been proposed for teleoperation, for example to reduce the velocity before impact (and the impact force) [19].…”