spot dog construction AI robot

Spot Robot Dog Spotted in Redwood City

The Northern California Healthcare Division and Innovations Team implemented Spot Dog to the test at a Medical Office Building in Redwood City, CA. In partnership with Avvir and Kaarta, the 197,800-square-foot medical center tracked progress on the 143 doctor offices and 116 exam rooms with technology solutions running parallel with the typical, more manual progress reporting. Swinerton partnered with Avvir to automate quality assurance through scanned-based AI and machine learning while Kaarta focused on Stencil scanning for our mobile mapping system.

During early scanning stages, a mobile scanner was used by physically holding the instrument upright while walking the jobsite.  These past few months, an automated method was introduced to test Spot Dog. At first, the robot was guided at a high level, but once a route for scanning was finalized, Spot was programmed with the route to run on its own as needed. The resulting real-time point clouds were optimized using the scanner’s onboard post-processing capabilities. The point cloud was transferred to the Avvir platform where it was compared to the BIM and previous project scans via a geometric analysis resulting in a construction progress report. Our project team was notified immediately through the online Avvir platform of installation deviations, project progress, and up-to-date installation quantities of materials on the project.  To ensure reporting metrics were accurate and in alignment with the client’s SOV format, Swinerton’s VDC team had several meetings with Avvir on best methods to extract and input metadata for each BIM element, such as, SOV ID, Material Type, System Type and Schedule Activity ID.

Before scanning, preliminary planning ensured a smooth process. The MOB site was scanned by each floor, with the scan path planned out in advanced to ensure each scan would be as efficient and effective as possible. Mobile scanning technology, with its real-time position estimation and rapid capture of data within a range of 300 feet, made for quick and easy scanning of the massive floor space.

According to Eric Law, our partnership with Avvir, Kaarta, and Boston Dynamics have been tremendous in researching automated data capture for validating quantity and quality of work put in place. This collaborative effort between our Innovations Team and our Northern California Healthcare Division has demonstrated great value to the project by providing data in generating payment applications, updating schedules, and cost forecasts.