Can AI Help Close the Timber Loop?

Demolition activities in the U.S. create 37 million tons of wood debris each year, with the vast majority ending up in landfill, according to a 2018 report from the U.S. Environmental Protection Agency (EPA). Harvesting and processing virgin timber can be energy and water intensive, can cause forest degradation, and comes with unknown levels of emissions that some argue are comparable to emissions from concrete or steel. And disposing of it leads to further emissions as decomposition releases the carbon stored in the wood. “Circularity”— achieved by salvaging and reusing old wood for new construction—is an obvious solution. Reclaimed wood keeps its carbon sequestered, avoids logging, and shortens the supply chain for new construction when processed locally—plus, using it in construction projects can satisfy green building requirements, like contributing toward LEED’s Sourcing of Raw Materials credit and the Responsible Materials and Responsible Sourcing imperatives of the Living Building Challenge. But using reclaimed wood requires the removal of all nails, screws, and other metal fasteners, which is time and labor intensive, and drives up costs.

Innovations in technology might be changing this. Urban Machine, a startup out of Oakland, California, developed a portable, modular machine that uses AI and robotics to remove nails and other metal fasteners from salvaged wood at scale. At 80 feet long, the machine sits on two 40-foot trailers, where it requires 6,000 ft² of space on jobsites and seven or eight operators. “We aim to be able to do 16,000 board feet a day. In relative terms, that’s a single-family house,” explains Jorie Wisnefski, marketing manager at Urban Machine.