Treeswift :: Robotics, Computer Vision & Machine Learning to Augment Field Workers Rather Than Replace Them

Treeswift is a Philadelphia-based company developing physical AI systems that combine robotics, computer vision, and machine learning to augment field workers rather than replace them. The company focuses on helping utility crews and other outdoor labor teams increase productivity in demanding environments such as power line inspection, vegetation management, asset documentation, and disaster response. Founded in 2020 by Steven Chen (CEO), Michael Shomin, Elizabeth Hunter, and Vaibhav Arcot, Treeswift has roots in university robotics labs including the University of Pennsylvania’s GRASP Laboratory and Carnegie Mellon’s Robotics Institute. The team blends deep expertise in robot perception, autonomous systems, and enterprise software, with members coming from Palantir, and other technology firms.

The company’s technology centers on sensor platforms that collect high-resolution data during normal field patrols, whether crews are walking or driving. Its primary robot-based products are wearable backpack-mounted sensor systems and vehicle-mounted sensor units. These platforms carry LiDAR scanners for precise distance and encroachment measurements, 360-degree cameras for comprehensive imagery, GPS for location accuracy, and microphones that use AI to transcribe spoken field notes automatically.

The data is processed through Treeswift’s AI software platform, which generates actionable analytics and a 360-degree “digital twin” of field conditions. This creates a virtual reality-style view of assets and surroundings that crews and planners can review remotely. The system integrates directly with existing utility GIS and planning tools so no new workflows are required.

Early versions of the technology, known as SwiftCruise, used autonomous drones capable of flying underneath forest canopies to collect tree-by-tree data for forestry applications such as carbon estimation, timber valuation, and growth forecasting. The company has since shifted focus primarily to electric utilities while retaining its robotics heritage. Current deployments support both routine (“blue sky”) operations and emergency (“black sky”) scenarios, including wildfire risk reduction, post-storm assessment, construction oversight, and vegetation management.

Treeswift’s platforms deliver measurable improvements by cutting the time crews spend on repetitive tasks such as manual measuring, counting, and data entry. Reports from utilities indicate significant efficiency gains, including lower fuel use during patrols, reduced truck rolls, faster identification of hazardous trees, and better prioritization of maintenance work. Major clients include Pacific Gas and Electric (PG&E) and several other large U.S. utilities. By giving crews richer, standardized field data without additional boots on the ground, Treeswift aims to make critical infrastructure work faster, safer, and more cost-effective.