The global site of the UK's leading magazine for automation, motion engineering and power transmission
28 March, 2024

LinkedIn
Twitter
Twitter link

Self-teaching robot picks random items with 94% accuracy

01 March, 2021

Automation experts at the UK’s Manufacturing Technology Centre have developed a robot that can teach itself how to pick up unknown objects from a randomly-arranged tray or bin. They say that tests on a wide range of objects – including metal components, cosmetic containers and fruit – have achieved success rates of up to a 94% for attempted picks.

The robot can pick objects without needing expensive sensors or lengthy programming. The technology – codenamed Project Viper – uses a deep neural network, low-cost depth-measuring cameras and robot arms fitted with vacuum pick-up cups.

“This project has shown the MTC’s ability to draw upon state-of-the-art academic research to develop new technologies for UK industry,” explains MTC senior research engineer, Mark Robson. “The result of this work is a flexible system that can quickly be adapted to the individual needs of different industrial handling problems without the need for complex robot programming.

“We showed that learning to pick on one object set can translate well to new objects,” he adds, “giving us an excellent starting point for new tasks. Building the system around a neural network architecture allows us to update the model as we gather new data in operation, enabling the system's performance to continue to improve over time.

“Using simulation to automate the creation of training data significantly reduces the cost and time typically required to manually produce the large quantities of data needed to train a neural network.”

The researchers at the Coventry-based MTC believe that the technology could bring big benefits to the manufacturing, agri-food, logistics and waste management sectors, among others.

The MTC’s self-teaching robot system uses vacuum cups to pick up items from randomly-arranged bins or trays

The picking technology follows an earlier MTC project that developed a robot with decision-making capabilities for assembly operations. Using a combination of machine learning and visual recognition, this robot can be taught to make assembly decisions based on the components put in front of it. The MTC says that it could save manufacturers the costs of expensive fixed tooling.

The MTC was founded by the universities of Birmingham, Loughborough and Nottingham, together with TWI. Its industrial members include major UK global manufacturers.

The MTC aims to provide a competitive environment to bridge the gap between university-based research and the development of innovative manufacturing systems, in line with the Government’s manufacturing strategy. It is part of the High Value Manufacturing Catapult, supported by Innovate UK.

MTC:  Twitter   LinkedIn




Magazine
  • To view a digital copy of the latest issue of Drives & Controls, click here.

    To visit the digital library of past issues, click here

    To subscribe to the magazine, click here

     

Poll

"Do you think that robots create or destroy jobs?"

Newsletter
Newsletter

Events

Most Read Articles