According to Bloomberg, SoftBank is backing Gravis Robotics with $200 million to retrofit existing excavators and earthmoving equipment with autonomous and semi-autonomous systems. CEO Ryan Luke Johns framed the investment around physical AI —retrofitting machines already on job sites for roads, quarries, and mines rather than building robots from scratch.
Construction firms don't buy sci-fi theater. They buy uptime, labor relief, and safety. One upgraded excavator that runs 24/7 with fewer operators matters far more to a contractor's margin than another chatbot feature. Serious AI capital is flowing toward solving labor scarcity in the physical economy, not chasing software-only returns.
SoftBank's Gravis bet reveals which AI track actually makes money. Autonomous construction equipment will outpace chatbot increments because labor is tight, equipment utilization drives profit, and a contractor's ROI on an autonomous excavator is measurable in weeks, not quarters. The next wave of AI winners will be the ones who attach autonomy to machines that already exist on job sites.
Filed to the Technology desk · Aug 19