- Register


Home>AUTOMATION>Handling & Logistics>Covariant launches to bring universal AI to robots

Covariant launches to bring universal AI to robots

10 February 2020

AI robotics company, Covariant, has launched from stealth announcing that its artificial intelligence has been deployed and is fully operational at customer facilities in the apparel, pharmaceutical and electronics industries.

Founded in 2017, Covariant has spent the past two and a half years researching, developing, testing and deploying its AI. Based on research in deep imitation learning, deep reinforcement learning and meta-learning, the company has pioneered the Covariant Brain: universal AI for robots that can be applied to any use case or customer environment. Covariant robots learn general abilities such as robust 3D perception, physical affordances of objects, few-shot learning and real-time motion planning, which enables them to quickly learn to manipulate objects without being told what to do.

Over the past year, Covariant graduated from its development and pilot phase into full production mode. Today, Covariant's robotic stations are running consistently at facilities in North America and Europe, including at a warehouse operated by Obeta (https://www.obeta.de), a German electrical supply wholesaler located outside Berlin.

"Customer expectations for fast, affordable package delivery have never been higher," said Michael Pultke, Head of Logistics at Obeta. "To stay competitive, we need to modernize our operations and keep order processing and delivery running quickly and smoothly. The Covariant-powered robot is an integral part of our live operations, exceeding our performance requirements and adapting quickly to change. AI Robotics is a foundational part of our future strategy."

Covariant brought Obeta's station into production in collaboration with KNAPP, a leading warehouse logistics technology company headquartered in Graz, Austria. Working with customers all over the world, KNAPP provides automation technology and software solutions for logistics and production facilities in the healthcare, textiles and fashion, trade, e-commerce and omnichannel retail, food retail and production industries.

"AI Robotics is quickly becoming relevant to every warehouse operation, and is critical to KNAPP's strategy of offering market leading solutions," said Peter Puchwein, VP of Innovation at KNAPP. "We looked at every solution out there. Covariant is the only one that's ready for real production."

Reaching continuous production is a big milestone for AI. The real world is notoriously challenging for robots because there's an unlimited number of scenarios that a robot could encounter, so it's impossible to program each one. In order to successfully operate and add value to customer environments, a robot must be able to see, understand, make decisions, learn from mistakes and adapt to change, on its own. Covariant's robots can do all these things continuously in real-world environments, powered by the Covariant Brain.

"At Covariant, we are dedicated to advancing research in AI robotics to solve real-world needs," said Peter Chen, founder and CEO of Covariant.

"Even though we are just getting started, the systems we have deployed in Europe and North America are already learning from one another and improving every day," said Pieter Abbeel, founder, President and Chief Scientist.

"The pace of research in deep reinforcement learning over the past five years has been extraordinary," said Geoffrey Hinton, Turing Award Winner, Engineering Fellow at Google, Professor of Computer Science at the University of Toronto, and Covariant investor. "Covariant has produced one of the first major breakthroughs in real-world robotic application, and it's just the start of the possibilities."

"Covariant has assembled an impressive group of roboticists and AI engineers," said Yann LeCun, Turing Award Winner, Chief AI Scientist at Facebook, Professor at NYU, and a Covariant investor. "If anyone can make robots learn to perceive, reason, and act in real industrial settings, it's Pieter, Peter and the Covariant team."

"Developing safe and reliable robotic technologies to address challenges in the real world is extremely hard," said Raquel Urtasun, Chief Scientist at Uber Advanced Technologies Group and Covariant investor. "Whether you're talking about self-driving cars or warehouse operations, robots encounter an endless number of unexpected scenarios. Covariant has demonstrated exceptional progress on enabling robots to fill orders in warehouses, which could unlock many other robotic manipulation tasks in other industries."

Covariant is backed by lead investor Amplify Partners and some of the world's top AI luminaries, including Jeff Dean, Geoffrey Hinton, Yann LeCun and Raquel Urtasun. To date, Covariant has raised $27 million in funding, including $7 million in seed funding raised in 2017 and $20 million in Series A funding raised in the first half of 2019, which had not been announced until today.

"Intelligent AI-powered robots have been science fiction mainstays for decades," said Sunil Dhaliwal, General Partner at Amplify Partners. "Covariant is the first company to make that dream a reality.  We have not seen another company with the ambition, technology edge, customer traction and raw talent required to bring AI Robotics to the physical world."

Covariant's team has decades of experience in artificial intelligence, deep learning and robotics. Pieter Abbeel's lab at UC Berkeley has pioneered many recent breakthroughs in robot learning, including a robot that organizes laundry, robots that learn (simulated) locomotion, robots that learns vision-based manipulation from their own trial and error and from human VR teleop. Before founding Covariant, founders Pieter Abbeel, Peter Chen and Rocky Duan spent over a year at OpenAI, pushing the frontier in deep imitation learning, reinforcement learning, unsupervised learning, and learning-to-learn.