A 1966 Ford Mustang has been converted into an electric vehicle that runs Tesla’s “Full Self-Driving” (Supervised) software. The project, completed by Calimotive Auto Recycling in Rancho Cordova, California, cost roughly $40,000 and took two years. It is likely the first non-Tesla vehicle to operate FSD.
Overview
Yaro Shcherbanyuk, owner of Calimotive, found the 1966 Mustang on Facebook Marketplace in summer 2022. He worked on the build with his father Viktor and brother Daniel. The team initially considered a Model S drivetrain but switched to a Model 3 setup after stripping the car. They grafted three sections of a 2024 Tesla Model 3’s floor and seats into the Mustang’s body, shortening the battery case to fit without altering the car’s original dimensions.
The result is a classic Mustang shell on a Model 3 dual-motor drivetrain producing roughly 400 horsepower and 471 lb-ft of torque. The car accelerates from 0-60 mph in about 3.5 seconds.
What the conversion includes
- Battery: 75kWh Tesla Model S battery (the source states a 75kWh battery, though the drivetrain is Model 3; the article is ambiguous on exact pack capacity)
- Drivetrain: Model 3 dual-motor setup
- Software: Tesla’s camera array retrofitted onto the Mustang, enabling Autopilot, Sentry Mode, and “Full Self-Driving” (Supervised)
- Interior: Model 3’s 15-inch touchscreen, Cybertruck yoke steering wheel, Tesla heated and cooled seats
- Charging: Tesla charging port located where the original gas cap was
- Efficiency: 258 Wh/mi — matching or beating a standard Model 3
- Range: 194 miles remaining at approximately 80% battery during a test drive with Business Insider
How FSD works on a non-Tesla
The most technically challenging part was getting FSD to function with camera angles different from Tesla’s production vehicles. Tesla’s vision-based neural network was trained on data from cameras mounted in specific positions. The 1966 Mustang has a completely different body shape, roofline, and mounting surface geometry — every camera sits at a different angle and height than the system was designed for. The fact that FSD still works suggests the neural net is more adaptable to non-standard camera placements than previously assumed.
Tradeoffs
- **Cost