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

Twitter link

ABB forms AI partnership to create e-commerce robots

12 March, 2020

ABB has formed a partnership with an American AI (artificial intelligence) start-up called Covariant, to bring AI-enabled robotics technologies to market, starting with autonomous systems for fulfilling warehouse orders. Their joint vision is to develop and offer AI-enabled robots that will work alongside humans in dynamic environments, learning and improving as they perform their tasks.

In particular, ABB has identified what it believes is a significant opportunity for AI-enabled robotics in applications such as logistics, warehousing, and parcel and mail sorting. The market analyst Statista has predicted that global revenues in this sector will climb by more than 50% from €1.7 trillion in 2019 to €2.6 trillion by 2024.

ABB believes that the growing demand for e-commerce fulfilment services, and that the labour-intensive nature of these businesses offers substantial potential for intelligent automation. The sector struggles to find and retain employees for picking and packing duties. While robots are ideal for repetitive tasks, they have, until now, lacked the intelligence to identify and handle the tens of thousands of constantly changing products encountered in typical dynamic warehouse operations.

Last year, ABB launched a global competition to assess how 20 AI technology start-ups would solve 26 real-world picking, packing and sorting challenges. The aim was to determine whether AI was mature enough to unlock the potential for robotics and automation in the segment. Covariant was the only contender whose software could recognise different items without needing any form of human intervention.

ABB was also looking for a technology partner with which to co-develop AI systems capable of supporting autonomous materials-handling, enabling its robots to handle a wide variety of items.

Covariant’s Brain AI software allows robots to see, reason and act in the real world, completing tasks that would be too complex and varied for traditionally programmed robots. The software enables the robots to adapt to new tasks through trial and error, thus broadening the range of objects that they can pick.

“Our partnership with Covariant is part of our strategy to expand into new growth sectors such as distribution and e-commerce and to leverage the scaling potential in these fields,” explains Sami Atiya, president of ABB’s Robotics and Discrete Automation business. “It perfectly complements our offering and adds to our aim to be the number one choice in robotics solutions for our customers. Through the combination of AI with our robots, we are opening an entire new field of opportunities and applications for a variety of industries.”

Covariant’s chief executive and co-founder, Peter Chen, adds that “we now need to take [AI] out of the laboratory and apply it to the real world”.

Applying Covariant's AI technology to ABB's robots allows them to adapt to new tasks through trial and error, and to handle a wide variety of items.

The combined ABB and Covariant AI technology has already been adopted by a Dutch company, Active Ants, which provides e-commerce fulfilment services for Web businesses.

Covariant is also working with an Austrian supplier of intralogistics systems, Knapp, to deploy and market advanced AI robotics technologies. Together, they have already developed several systems, including a robot designed for high-performance single-piece picking applications. Called the Pick-It-Easy robot, the machine is already in use at several sites in North America and Europe, including at a German electrical wholesaler, Obeta, at a site near Berlin.

“Artificial intelligence will be a defining feature of the warehouse of the future, impacting all aspects of operations and fundamentally changing how business is done,” says Jusuf Buzimkic, Knapp’s senior vice-president of engineering. “Our partnership with Covariant will enable us to deliver cutting-edge AI technology to our customers, providing a major leg-up in an increasingly competitive world.”

He adds that the Pick-It-Easy robot can handle unlimited SKU types and challenging objects such as polybags, transparent objects and blister packs. “It also learns to pick new objects it’s never seen before and improves over time,” Buzimkic reports.

Covariant, based in Berkeley, California, was founded in 2017 by AI researchers and roboticists from UC Berkeley and OpenAI. It has already raised $27m in funding.

  • 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



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



Most Read Articles