Principal investigator: Kaspar ALTHOEFER
Funding source(s): Innovate UK
 Start: 01-10-2017  /  End: 31-03-2019
 Amount: £97,279
Directly incurred staff: Jelizaveta Konstantinova

In this project, we aim to create an autonomous grasping system, based on existing technology developed by the Shadow Robotics Company (SRC) and sensor experts at The Centre for Advanced Robotics @ Queen Mary (ARQ).

The proposed project involves the feedback from prospective industrial customers at every stage of the project. This work builds on global research into grasp planning algorithms, and the UK sensor expertise at QMUL to translate the technology from research to commercial application across sectors including domestic service robotics, factory logistics, flexible automation, and pharmaceutical lab analytics. In each of these sectors we have identified customers who can deploy thegrasping intelligent grasping capability.

The robotics team at QMUL will contribute in the areas of tactile, force and proximity sensing. In addition QMUL will focus on expanding the created sensor fusion algorithms to enhance the flexibility of grasping, grasp correction and manipulation tasks across different objects. 

Sensors will improve the reliability of the new hand prototype with regards to achieving good grasp closure duringmanipulation tasks and extend the grasping capabilities of the hand for a wide range of objects. Further, we will focus onthe development of a multi-modal sensor sensing system that is capable of extracting relevant features from raw sensorsignals and providing a combined interpretation of sensor information to be used for grasp adjustment and grasp qualityassessment.  For the duration of the entire project, we will be ensuring that the work conducted meets the needs of industrial customers and that those customers are engaged with the development process throughout. Therefore, the sensor design
and classification algorithm will be refined iteratively during the development stages of the project with any advances being fed into the benchmarking of the system.