Constructor Autonomous Racing

AI driving at the limit of the real world

Professional AI racing team competing in A2RL, the world’s leading autonomous racing leagueFormula 1 for AI

Identical race cars
No human drivers
The winner is the software

8years in autonomous racing
1,000hours on track and test bench
27patents in autonomy and robotics
15+ TBof proprietary racing data

Team

Professional engineering team, not an academic project

The team is built by experts in AI, robotics and dynamic systems engineering, and designs, tests and races the full autonomy stack: perception, planning and control. We are based in Germany and represent Constructor Labs and Constructor University.

2022Roborace World Champions
20242nd place, A2RL Grand Final
20262nd place at A2RL Imola, Europe’s first multi-car autonomous race

Constructor Racing’s autonomous Dallara Super Formula car, number 8, in side view

Record

Eight years at the front of autonomous racing

2018
The team is formed
2019
The team joins Constructor
2020
Roborace founding team, multiple race wins
2021
Leads most of the Roborace season
2022
Roborace World Champions
2023
Joins A2RL
2024
2nd, A2RL Grand Final
2025
6th, A2RL Gold Final
2026
2nd at A2RL Imola, Europe’s first multi-car autonomous race

5 podiums

League

A2RL: Formula 1 for AI

The world’s leading professional autonomous racing league. Every team runs the same car. The whole competition is in the software: perception, planning and control.

Visit a2rl.io

The full A2RL grid of autonomous race cars and teams on the start straight at Imola — The A2RL grid at Imola, Europe’s first multi-car autonomous race

280km/h

Top speed of the car, with no one on board

Perception sensors on every car: 3 LiDARs, 4 radars, 7 cameras

Teams in the league

Drivers and remote controls. The AI makes every decision

Identical full-scale Dallara Super Formula car for every team
Wheel-to-wheel, multi-car racing, not time trials
A2RL founded in Abu Dhabi; first fully autonomous race in 2024
A2RL races at the Formula 1 Yas Marina Circuit in Abu Dhabi
First European A2RL race at Imola, September 2026
Three A2RL races in 2027: China, Europe and the UAE

Technology

Physical AI that acts in the real world

AI controls real race cars, in real time, wheel-to-wheel at high speed. We test every software release on a real race car at up to 280 km/h before it goes into a race.

01 — PERCEPTION

See

Understanding the track, the car and every rival around it.

Computer vision
Sensor fusion: LiDAR, radar, cameras

02 — PLANNING

Decide

Choosing the fastest safe line, overtakes and defence, many times a second.

Machine learning
Optimised path planning and prediction

03 — CONTROL

Act

Turning decisions into steering, throttle and brake at the limit of grip.

Vehicle dynamics
Safety-critical control systems

How we improve Data→Model→Track→Improvement↻

Constructor Racing’s autonomous car on track at Imola

Why it matters

If AI can operate safely at racing speeds, it can power the next generation of autonomy across industries.

Autonomous racing is the ultimate real-world benchmark for Physical AI. It validates perception, decision-making and control under pressure, at speed and in the face of risk. This is where AI moves beyond simulation and proves itself in reality.

Beyond racing

Racing is our testing ground for real-world deployment

The same stack that races at 280 km/h is built into products and services for motorsport, mobility and industry.

Racing Coach

A real-time AI driving coach: a lap-by-lap plan, racing-line corrections and throttle and brake deltas against the AI-optimal lap.

For racing schools, OEM driver programmes, sim racing and track days

ADAS & obstacle avoidance

High-speed collision avoidance for road and private-area vehicles, built on our racing perception and planning.

For carmakers and vehicle programmes

Private-area autonomy

Self-driving for controlled areas where the racing stack transfers directly: ports, yards, campuses, agriculture and logistics sites.

For ports, logistics, agriculture and industrial sites

Autonomy kits

Modular perception, planning and control kits for target industries and fleet orchestration.

For ADAS and autonomy teams

R&D consulting & validation

R&D in ADAS and private-area autonomy, plus extreme-edge sensor and compute validation on our simulation, telemetry and test bed.

For OEMs, Tier-1 suppliers and deep-tech investors

Have a problem the racing stack could solve?

Tell us about it. Our engineers will come back to you.

Trusted by

Our racing technology is used by

The only autonomous racing team whose clients include classic motorsport teams.

Audience

Where the industry watches

Three years old, with an audience growing by double digits with every race. Autonomous racing draws carmakers, engineers, investors and the world’s press.

A2RL Imola, 5 September 2026

150+

VIP guests at the networking event

~60

companies represented

35+

journalists at the media day

80+

articles published

A2RL reach, per race

1.1B

households reached

147M

social impressions

YouTube race views

livestream views

Covered by

Figures: A2RL partner report, Imola, September 2026.

Our partners

Racing together

The companies that back our team on and off the track.

Constructor Racing car number 8 in Corpay livery racing at Yas Marina

Team partner

Corpay is a global corporate payments company and a member of the S&P 500 (NYSE: CPAY). It helps more than 800,000 businesses pay expenses in a simple, controlled way, from supplier and cross-border payments to fuel, EV charging, tolls and workforce lodging.

Visit corpay.com

Partner with us

Join the future of Physical AI

Position your company in Physical AI with a team that proves it on track.

Backed by Constructor

Science, engineering and business behind every lap

Constructor Racing is part of Constructor Group, a private education, research and technology group. A rare combination in the field: a private university and a private racing team.

Research and talent

Constructor Knowledge brings together education, research and talent development across an international network, including Constructor University in Bremen, Germany (2,500 students on campus), EBAC in Brazil and Nexford in the United States.

constructor.org/knowledge

Research into real-world solutions

Constructor Labs bridges advanced scientific research and real-world applications, turning cutting-edge technologies into practical solutions. Its expertise spans AI, robotics, autonomous systems and data-driven innovation, connecting research with industry needs.

constructorlabs.org