# Learning Interactive Behaviors for Musculoskeletal Robots Using Bayesian Interaction Primitives

## TL;DR

This paper introduces a Bayesian Interaction Primitives approach enabling musculoskeletal robots to learn and respond interactively in real-time, despite their complex nonlinear dynamics and limited modeling, demonstrated through handshake tasks.

## Contribution

The paper presents a novel method for learning interactive behaviors in musculoskeletal robots using Bayesian Interaction Primitives from limited demonstrations.

## Key findings

- Capable of real-time state estimation and response generation.
- Generalizes to new positions, partners, and velocities.
- Effective for human-robot handshake interactions.

## Abstract

Musculoskeletal robots that are based on pneumatic actuation have a variety of properties, such as compliance and back-drivability, that render them particularly appealing for human-robot collaboration. However, programming interactive and responsive behaviors for such systems is extremely challenging due to the nonlinearity and uncertainty inherent to their control. In this paper, we propose an approach for learning Bayesian Interaction Primitives for musculoskeletal robots given a limited set of example demonstrations. We show that this approach is capable of real-time state estimation and response generation for interaction with a robot for which no analytical model exists. Human-robot interaction experiments on a 'handshake' task show that the approach generalizes to new positions, interaction partners, and movement velocities.