Ddespite robotics’ huge promise to change life as we know it, today’s robots are largely dumb and inflexible as they grapple with a complex and ever-changing world. “No one was doing it right, and it would be huge if we could do a better job,” says cognitive science researcher Gary Marcus.

So Marcus, a former NYU professor and founder of another startup called Geometric Intelligence which was acquired by Uber, teamed up with roboticist Rodney Brooks, co-founder of iRobot and co-inventor of Roomba, to create a new business. Their startup based in Palo Alto, California, Robust.IA hopes to bridge the gap between what robots can do today and what they promise by building what they call “the world’s first industrial-grade cognitive engine for robotics.” Its goal is to help companies quickly build and deploy semantically conscious robots that could eventually work on construction sites, care for the elderly, make fully autonomous driving a reality, or perform other complex tasks.

“We want to build a cognitive engine that will allow people to configure robots faster and have more confidence in their reliability,” said Marcus. “Instead of hiring 40 doctors and taking three years, they should be able to take six months with maybe five doctors. It’s the vision of what we’re trying to build, it’s a long-term vision.

Operating largely under the radar since its founding in 2019, Robust.AI said today it has raised $ 15 million led by Jazz Venture Partners and joined by former investor Playground Global, bringing its total funding to $ 22.5 million. dollars. Marcus, the CEO of the company, declined to disclose the company’s valuation other than to say that “it’s not obscene, but it’s healthy.” Other investors in the company include angel investor Esther Dyson, Veritas software co-founder Mark Leslie, Skype co-founder Jaan Tallinn and Silicon Valley entrepreneur Steve Blank.

With this new funding, Robust.AI, which has some 25 employees, intends to develop the software, which focuses primarily on social perception and navigation, and get sellers to start marketing its offerings. Marcus says the company is currently running a pilot project with a client he refuses to name and is “in conversation with several other partners.” As Robust.AI is a software company, its partners will need to provide the hardware. Marcus hopes to start shipping products in 2021. Building a business like this is a long and difficult process, and even if everything goes according to plan, it will be time before his software is widely available under license.

“We want to build a cognitive engine that will allow people to configure robots faster and have more confidence in their reliability,” says Gary Marcus of Robust.AI.

The company’s roots can be traced back to conversations Marcus, 50, and Brooks, 65, began to have about issues around artificial intelligence and its application. “We believe there is a lot more to AI than just big data sets,” says Brooks. “I am of the opinion that it is almost a fetish.”

The two connected after Marcus was working on his latest book, AI restart, co-authored with NYU computer science professor Ernest Davis, on the limits of artificial intelligence. He sent the chapter on robotics, which explored the gap between robotics’ potential and its reality, to Brooks. “He said to me, ‘You’re way too nice,’ which people don’t tell me,” recalls Marcus.

Research on the book had also sparked the idea for the company, and he offered Brooks to become a consultant. “It was kind of a flyer. I was impressed with what he had accomplished, ”he says. “Then I spent three months recruiting him. “

At the time, Brooks was terminating collaborative robotics startup Rethink Robotics, which closed in October 2018. He quickly signed on as co-founder and chief technology officer. The company’s three other co-founders have similar stellar pedigrees: Mohamed Amer previously led AI and machine learning projects at SRI International; Anthony Jules had been CTO of Formant.io, an intelligence platform for robot fleets, and senior product manager at X; and Henrik Christensen is a professor at UC San Diego and director of its Contextual Robotics Institute.

“We wanted to get a software platform that would help robots work in the real world, and not in these constrained environments,” says Marcus. “There’s nothing on the shelf where if you want to build a household robot, you can start as much as you can if you want to get a game engine and make a video game.”

With a game engine, he explains, video game developers can let the software solve time-consuming problems, like where objects will collide, and focus on the creative aspects of the game. he thought, should have the same base, rather than requiring engineers to hack together a mix of robot operating system, or ROS, and other programs tailored to specific problems.

As the use of artificial intelligence and automation expands into areas that were not previously there, the need for such foundational software is expected to increase. “A lot of businesses that haven’t previously thought of AI in their systems will need or want it,” says Brooks. “Not all of them can have world-class AI research labs. It must be a commodity for them.

In the short term, says Marcus, Robust.AI focuses on social browsing, which is the simple problem of getting around people or objects without interrupting or being intrusive. Longer term, he has ideas on how to handle language more accurately and efficiently so that, for example, a request to put things in a closet does not result in the hacking of those items by the robot to adapt them.

“We wrote this great chapter [in Rebooting AI] why the language approach taken by deep learning would never work, and I stand by that assertion, ”he says. “The fundamental problem is that we don’t say everything we need to. “