Daily household tasks involve manipulation in cluttered and unpredictable environments and service robots require complex skills and adaptability to perform such tasks. To this end, we developed a teleoperated online learning approach with a novel skill refinement method, where the operator can make refinements to the initially trained skill by a haptic device. After a refined trajectory is formed, it is used to update a probabilistic trajectory model conditioned to the environment state. Therefore, the initial model can be adapted when unknown variations occur and the method is able to deal with different object positions and initial robot poses. This enables human operators to remotely correct or teach complex robotic manipulation skills. Such an approach can help to alleviate shortages of caretakers in elderly care and reduce travel time between homes of different elderly to reprogram the service robots whenever they get stuck. We performed a human factors experiment on 18 participants teaching a service robot how to empty a dishwasher, which is a common daily household task performed by caregivers. We compared the developed method against three other methods. The results show that the proposed method performs better in terms of how much time it takes to successfully adapt a model and in terms of the perceived workload.

Certain tele-manipulation tasks require manipulation by two asymmetric slaves, for example a crane for hoisting and a dexterous robotic arm for fine manipulation. It is unclear how to best design human-in-the-loop control over two asymmetric slaves. The goal of this paper is to quantitatively compare the standard approach of two co-operating operators that each control a single subtask, to a single operator performing bimanual control over the two subtasks, and a uni-manual control approach.

In tele-operation, haptic feedback from the remote environment to the human is often limited, which has been shown to negatively influence the performance and required time of tasks. The conventional research focus is on improving the quality of the haptic feedback (transparency), which may have led to significant improvement, but is still imperfect, with many unresolved issues. The present study presents an alternative approach to improve tele-operated tasks: by offering haptic shared control in which both operator and support system apply the required forces at the input (master) device. It is hypothesized that virtual forces from well-designed shared control will improve required time and accuracy, with less control effort, and that these benefits exist for perfect transparency but even more so for imperfect transparency. In an experimental study haptic shared control was designed to aid operators (n=9) with performing a simple bolt-spanner task using a planar (2D, 3DOF) tele-operator setup. The experimental results provided evidence for the hypotheses, showing that the tested tele-manipulation task benefits from haptic shared control, for three different levels of transparency. Essentially, the presence of haptic shared control allows for a worse transparency without compromising required time, and can even improve required time during perfect transparency.