Please turn your device to portrait.
Urban Machine pairs AI and robotics to save timber from landfill with opportunities for adaptive reuse and prefabrication.
A U.S-based startup in Oakland, Urban Machine, is seeking to address two major issues in the construction industry, namely waste and labour shortage, through the use of artificial intelligence and robotics.
The company is currently developing a large robotic disassembly line that can automatically remove metal bits, such as staples, nails, and screws, from timber collected from construction and demolition sites. Each machine weighs over 22 tons and is approximately 24 metre long. They’re aiming to process at least 4,800 metres of timber daily.
The company has built three prototypes, and the next one will be portable enough to be transported to construction and demolition sites via truck. The machines use computer vision to identify the type of metal pieces that need to be removed, and AI-powered algorithms help identify the type of fastener present, after which a robotic arm with manipulators pulls or screws the metal out.
Urban Machine currently has around a dozen employees, and plans to deploy a dozen of its machines across six U.S cities in 2024. Each machine requires five or six people to operate, and the company estimates a need for 6,000 machines around the country to reclaim half of all timber that would otherwise be wasted.
The plan is to sell the processed timber to commercial manufacturing companies for reuse in prefabrication projects that can handle the type of dense, hardwood their tech will reclaim.
The company’s goal is to sell or lease the machines with software and maintain rental fleets.
Watch a machine in action
Urban Machine has raised AUD $11 million through a seed round from investors such as GV, Lowercarbon Capital, Catapult Ventures, Klein Venture Partners, and Union Labs Ventures. Law.
According to a report by Grand View Research, globally, the market for reclaimed timber is expected to grow 4.6% annually and exceed AUD $104 billion by 2028.
