Problem

Flywheel AI addresses the construction industry’s shortage of skilled equipment operators, high safety risk, and operational inefficiency. Its launch materials describe a U.S. shortage of roughly one million workers, more than 1,000 annual safety deaths, hundreds of thousands of dollars in penalties and multi-week shutdowns per incident, and up to 40% operational downtime from blind spots and inefficiencies. The work is particularly “dull, dirty, dangerous,” making excavator operation a compelling starting point.

The killer use case is allowing contractors to operate excavators remotely, reducing the need to place workers in hazardous environments while improving fleet utilization. A startup profile estimates approximately $150,000 in annual operator savings and $500,000 in avoided safety costs per machine, although those figures are company-oriented commercial estimates rather than independently verified financial results.

Product / Service

Flywheel AI offers a retrofit system that converts existing excavators and other hydraulic equipment for remote operation rather than requiring contractors to purchase specialized new machines. The system is described as machine-agnostic, installable in a few hours, and compatible with continued manual operation. Its low-latency teleoperation platform lets a supervisor remotely turn machines on or off and take over operations using a relatively minimal setup of a laptop and joysticks.

The product also creates a data and autonomy layer: teleoperated machines collect real-world operating data used to train models for tasks such as digging, dumping, trenching, and grading. This gives customers an immediate safety and productivity benefit from remote operation while creating a path toward workflow autonomy, in which one operator can oversee more machines and sites. The commercial model appears to combine retrofit hardware with recurring teleoperation and autonomy software subscriptions.

Market

Flywheel AI competes in construction robotics, heavy-equipment teleoperation, and supervised autonomy. Relevant competitors include Teleo, Built Robotics, Safe AI, and equipment manufacturers developing their own automation. Teleo is the closest described analogue because it also retrofits excavators and other heavy equipment for remote or semi-autonomous operation, while Built Robotics represents a more fully autonomous excavator approach. An external market profile frames the opportunity around a roughly $2 trillion global construction market, six million excavators, and a potential $180 billion retrofit market.

Flywheel appears to be early-commercial rather than a scaled company. Y Combinator lists it as active, founded in 2023, part of its Summer 2025 batch, and operated by a two-person team. A separate startup profile reports early revenue from three companies, a growing multimillion-dollar pipeline, and $500,000 of pre-seed funding, while a social-media report says its excavator AI co-pilot was being deployed on major construction sites in the United States and Singapore. Public evidence does not identify those customers or provide audited revenue, so the best-supported characterization is early traction with deployments and initial revenue—not an established, broadly scaled business.