Skip to content

SOFTGRIP: Functionalised soft robotic gripper for delicate produce harvesting powered by imitation learning-based control

The SOFTGRIP project will deliver an innovative, soft gripper robotic solution for the autonomous picking of delicate white button mushrooms cultivated on Dutch shelves.  The fresh food industry is highly labour-intensive with associated costs contributing up to 50% of overall production costs.  Therefore pressure is mounting to reduce these costs, at the same time as the industry is facing major labour shortages.  So far, robotic automation for picking delicate fresh produce has been impossible due, largely, to the complex, contact-rich interactions involved in such tasks. 

In seeking to address this the, SOFTGRIP consortium will develop: 

  • A low-cost, soft robotic gripper with built-in actuation, sensing and embodied intelligence that enables reliable and efficient mushroom picking
  • Material synthesis and fabrication techniques that offer precise tuning of mechanical properties, comply with food safety standards, allow for chemical recycling and offer self-repair properties
  • Machine vision tasks for the detection and identification of mushrooms; and estimation of the position and orientation of the soft gripper
  • A set of accelerated continuum mechanics modelling algorithms that facilitate real-time, model-based, control schemes, capable of being executed by limited computational resources
  • Advanced learning capabilities for the gripper, through a learning-by-imitation framework consisting of multi-task and meta-learning techniques, so it can be deployed with minimal programming

 On completion, SOFTGRIP will enable a step-change in efficiency, helping mushroom growers to reduce costs by >30% and increase yields by >20%, while also improving job quality in the industry.  In the longer term, this versatile solution will lower the barriers to robotics deployment and open up new opportunities for the adoption of robotic solutions in the agri-food sector, in particular with relation to other fresh foods which require similar stringent handling, such as kiwi fruit, grapes and berries.

Partners: Sant’Anna School of Advanced Studies, Institute of Communication and Computer Systems, NTUA, University of Essex, TEAGASC - Agriculture and Food Development Authority, MITSUI Chemicals Europe GmbH and the Essex Innovation Centre.

SOFTGRIP is funded by the European Union’s Horizon 2020 programme under the ICT call “Research and Innovation boosting promising robotics applications”.

Subscribe to our newsletter to receive the latest news and events from TWI Innovation Network:

Subscribe >