4 issued U.S. patents.62 claims.

An AI control system that scans a vehicle in 3D, chooses the right tool head for the surface in front of it, and carries out multi-step surface treatment — film application, wrapping, coating, tint — while detecting defects and correcting in real time.

• 3D vision using stereo cameras and laser grid projectors to build high-resolution surface models
• An automated magazine of interchangeable, AI-controlled tool heads
• Tool heads including a heated flexible spatula, a motorized film tensioning tool, a vibration-assisted squeegee, and knifeless tape for bladeless trimming
• Autonomous defect detection with real-time corrective action
• Adaptation to ambient temperature, humidity, wind and surface contamination
• Mapping human expert demonstration onto robotic joints through imitation learning

A system and a method for installing and servicing vehicle lighting — fiber-optic starlight headliners, LED strips, headlight and taillight assemblies — using robots that build a 3D model of the vehicle, plan the work, and place components with tactile feedback.

• A 3D digital model of the vehicle interior and exterior, built at fixed facilities or on-site at customer premises
• Compensation for variable ambient lighting, temperature and ground irregularities
• An installation plan generated from a desired lighting pattern or a diagnostic task
• Precision placement maintaining the integrity of headliner fabric, plastic trim and wiring harnesses
• Autonomous wiring routing and interfacing with vehicle diagnostics for functional verification
• Real-time quality control where findings automatically update the machine-learning models

An AI system for running a vehicle service facility: reading vehicle diagnostics, generating explainable repair strategies, dispatching work to robots with assigned roles, maintaining a digital twin of both the vehicle and the facility, and keeping customers informed.

• Multi-modal sensor data and vehicle diagnostics, including EV and software-defined vehicles from model year 2005 onwards
• Explainable-AI repair strategies
• A digital twin of both the vehicle and the service center
• Humanoid roles spanning check-in, payment, diagnostics, repair, logistics and security
• Force and tactile sensing for fine manipulation, including high-voltage EV components under specialized safety protocols
• Asset-utilization optimization across staging areas and bays, including triaging previously unfinished projects

A method for tracking how a pigment pattern in skin changes across treatment sessions — normalizing photographs taken under different lighting, quantifying the change, and forecasting a completion date with a confidence weighting.

• Normalizing images for environmental lighting using deformable registration against stable epidermal micro-landmarks
• Conditioning on capture metadata — focal length, exposure time, ISO, white balance, illumination intensity
• Quantifying pigment-pattern change with a machine-learning model
• Forecasting future change by non-linear regression, including auto-detected Fitzpatrick skin type
• Displaying a progress metric and a confidence-weighted completion date from a patient-specific curve