Urban Machine uses Zivid 3D to reclaim lumber and salvage the past to build the future

01 Story

Every year vast amounts of lumber are consumed by the construction industry. In 2019 the United States consumed 47 billion board feet of lumber. Whilst a large amount of this lumber is used as an integral part of a building structure, a very large portion is used for temporary purposes, such as shuttering for concrete and supporting purposes. Once construction is complete, much of this timber is deemed unusable and marked up as CDWW (construction demolition wood waste) and is typically cast after use and ends up in a landfill. The reason much of this lumber cannot be reused is because it has fastenings embedded in it, these include nails, staples, and screws. 

It simply isn’t feasible economically, on a time-taken-to-do basis, or to find people willing to do such strenuous and tedious work. This waste lumber is significant, in fact, some studies estimate it is approaching 40 million tons every year in the US alone. Urban Machine came about because the founders weren’t prepared to sit by and see such unnecessary wastage of a natural resource that was ripe for reuse. An additional driver is CO² emission reduction. Lumber and timber absorb and store carbon over time, it is estimated buildings in the US currently store over 7 billion tons of carbon. Reusing this timber means keeping that carbon trapped in the existing wood and not allowing its release, as would happen when scrapped. Urban Machine have developed ‘The Machine’ to help solve this challenge and strike a blow for a better, more sustainable world. 

02 Challenges

The Machine is a perfect fit for 3D machine vision. While the lumber is fed into the machine in standard dimensions by length and width, it can see any type and number of fastenings located in it. In addition to all the locations in the timber, the fastenings will never be in a way that can be predicted. The machine vision must be able to give a real-time picture of the state of each board. The fastenings are almost all small, and some extremely small. A typical nail might be 3 or 4 millimeters in diameter, but a small staple can be just a millimeter wide. The 3D machine vision must be able to deliver the most acute, accurate, and detailed point clouds for the robot to locate and remove effectively.