How many robot helpers are too many?

AI that one can follow seems like a simple enough task. Asking someone to do it is certainly a simple thing to do, but what if people or objects get in the way of the robot following behind a person? How do you manage an ever-changing environment?

About a year ago, at a TechCrunch robotics conference held at UC Berkeley, the founders of AI startups explored solutions to common problems that occurred when trying to automate construction projects. Dusty Robotics CEO Tessa Lau called attention to the challenge of moving machines in an unstructured environment full of people.

“[The] normal construction site – it is cumbersome, and anyone with knowledge of robotics who knows anything about robots knows that it is very difficult to get robots to work in the that kind of unstructured environment, ”she said.

That’s why this week Piaggio Fast Forward, maker of personal robotics Gita, shared details of work they are doing with industry technology services provider Trimble using the Boston Dynamics API to create robots that follow construction workers. Pilot tests were conducted at an office building under construction in Colorado.

Robots can also travel in groups with one Spot Mini and two Gita robots in what Piaggio Fast Forward calls platooning. As part of the construction pilot, Piaggio is evaluating people’s perceptions of how many robots are following human overgrowth.

Making the most of the number of robots that follow people on a construction site is a concern not only whether the robot can safely make it out of the way of a dozer led by man, but also the question of how many robots can be involved before strange things happen. So, if you’re a construction worker looking at a robotic site manager approach, are five robots the top line? Six? Piaggio Fast Forward CEO Greg Lynn told VentureBeat that the company could support a convoy of 50 to 100 robots but that this would be inconvenient.

“How long could a platoon of robots be accepted, just as a measure? It looks like it’s about 15 or 20 feet. To be completely honest, we don’t know yet, but it’s not that long, “he said.

The approach uses Trimble lasers and geospatial sensors to map the environment and establish space. One of Spot Mini’s experiments involved training a robot to walk a path once and then try to repeat that path automatically. The approach follows how people manage the landscape, understanding the movement of people rather than relying solely on mapping.

Piaggio Fast Forward introduced its Gita robot for tracking behind humans and carrying up to 40 pounds of weight in 2019. The machine can reach speeds of 22 mph. The Trimble collaboration is the first industrial use of Piaggio tracking technology, Lynn told VentureBeat. Piaggio Fast Forward is part of the Piaggio Group, one of the largest scooter and motorcycle manufacturers in Europe, best known for their Vespa brand.

Autonomous trucks have been operating in convoys before, but as part of a bid to enter more business applications, a spokesman told VentureBeat that next month the company will test platooning for grocery delivery in a planned community with several thousand homes.

As part of experiments announced in late 2020 without platooning, Gita robots are being used at Cincinnati and San Diego airports and, by Doğan Group, a shopping center in Turkey.

Learning how to build unstructured environments such as construction zones can lead to stronger AI that will help farmers in fields, delivery robots in cities, or people in wheelchairs on steep slopes. Should these systems be reliable and trustworthy, they could be common in business environments for surveillance – but going further, why push a wheelbarrow if you can just follow?

Ideally, a robot companion on a workstation could augment human workers by carrying devices for electronics, as the demo video below shows. They could also use computer vision to alert them to safety hazards or monitor the progress of construction projects. One of the earliest applications of Spot robots on construction sites was to inspect construction projects. Spot Mini has also been used in law enforcement by the New York Police Department, which is locating Spot’s position after a police officer walked a blow.

Piaggio is not alone in his goal of implementing people tracking technology in the business workplace.

VentureBeat covered a $ 10 million funding round for ForwardX, which was famous at the time for luggage carrying passengers. A company spokesman told VentureBeat that the company, which is based in China, was out of baggage during the pandemic and is now aiming for human search solutions for robots that work together. to people in warehouses and manufacturing facilities. To date, ForwardX has run test cases in factories for companies such as DHL and Toyota, among other customers.

As AI companies like Piaggio Fast Forward explored opportunities in the construction industry, traditional players like Caterpillar and Komatsu last year saw increased sales of autonomous software for business environments such as construction, mining and space.

Gita robots are not designed to move around in a landscape of scattered dirt where dozers and earthmovers work, but Lynn said once concrete slabs are laid for a construction project, Initial tests showed only a few areas that the robots could not travel.

“One of the things about the Trimble naming is that we are kind of announcing that we don’t have to build every robot on Earth. We want to engage with people who have problems with this because we are a software company as big as a hardware robotics company, ”said Lynn. “We do not want to do everything ourselves. And we don’t think it’s 100% of the solution anymore. ”

There are several entry-level AI stages in the industry of building automation. In 2019, Built Robotics spoke to VentureBeat about efforts to automate dozers and create forecasting systems for project management and other purposes. Last fall, Canvas emerged from stealth to introduce AI and robotics into construction zones in the San Francisco Bay Area to install drywall, and in January Swapp raised $ 7 million to compete among a fleet of businesses who worked to automate construction project design mapping.

The use of robots in construction environments has led to some accidents. According to safety and compliance website BLR, in 2019 two construction workers were injured in crashes involving robots demolished on construction sites in Washington state, with robots crushing one worker against a wall and crushing someone else’s leg.

In other news about robots navigating unstructured or industrial conditions, in May 2020, Burro, which develops robots that follow farm workers, helped harvest grape in Coachella, California. And last fall UC Berkeley AI researchers introduced LaND, an AI that learns from disconnection times to improve the delivery of delivery robots on trails.