RoVi - Robot Vision

Overview

We aim to develop a novel visuo-haptic sensor system for robotic manipulators, which allows for multimodal sensing, reduced system complexity and significantly lower costs compared to current systems. It replaces dedicated force sensors with passive components and a camera, providing coherent measurements of both force and contact shape. 

Problem

Intelligent robots rely on sophisticated perception systems for planning and manipulation in complex scenes. Many distributed sensor modules are required, such as cameras, force-torque and tactile sensors. This results in:

• High hardware complexity
• High assembly and maintenance costs
• Complex sensor calibration and synchronization

Technology

Our patented technology substitutes active sensor modules for force-torque and tactile data by an external camera and a passive flexible element. Forces applied to the passive element result in a characteristic deformation.

Robust image processing algorithms detect the deformation from the camera image, and provide the applied forces or contact patterns for further control and planning tasks.

Advantages

• Low cost – based on ordinary flexible materials
• Simple hardware – no electronic components and no wiring within the sensitive region
• Low complexity – hardware replaced by software
• Adaptability – fits actuators of any shape
• Compatibility – possible to extend existing systems
• Rich feedback – visual & tactile data from one sensor
• Versatility – vision tasks on the same camera