MenteeBot by Mobileye Global Inc.: status and timeline

Milestones

Next · Autonomous no-teleop proof-of-concept deployments

Mentee Robotics founded2022Complete.
Mobileye acquires Mentee RoboticsJan 2026Complete.
Autonomous no-teleop proof-of-concept deploymentsTarget 2026Current milestone.
Series production launchTarget 2028Not yet reached.

Most important updates

Mobileye acquires humanoid robotics startup Mentee Robotics for $900 millionJan 2026

Upcoming

2026Autonomous no-teleop proof-of-concept deployments (next)
2028Series production launch

Current obstacles

Reality gap in simulation transferBehaviors learned in simulation may fail in real robots due to unmodeled physics, sensor noise and environmental variability.
Complex task learning from limited demosLearning diverse industrial tasks from single or few human demonstrations remains a research frontier.

Physics limits

Demonstration quality matters greatlyIf a human demonstrates a task poorly or with unspoken assumptions, the robot learns those flaws.
Simulation fidelity must be highThe more accurate the simulator, the better the transfer; high-fidelity simulators are computationally expensive.

How it works

4 parts

Human demonstration

Learn from single examples

A human operator shows the robot a task once; the system learns the core behavior from this demonstration.

Simulation scaling

Millions of virtual training iterations

The learned behavior is transferred to a physics simulator where it runs millions of variations, learning generalizations without real-world data collection.

Real-world deployment

Transfer to physical robot

The trained policy is deployed on physical MenteeBot, which executes complex tasks autonomously in industrial settings.

Continuous improvement

Learning from deployment

Real-world experience feeds back into the simulator, improving the learned models for future deployments.

Spec sheet

Spec	MenteeBot
Training approach	Simulation + human demonstration generalization^R (reported)
Operation mode	Autonomous, no teleoperation required^R (reported)

^R reported by the company

Update log

1 update

Jan 2026

Major: Mobileye acquires humanoid robotics startup Mentee Robotics for $900 millionPressCommercial

About Mobileye Global Inc.

The team behind MenteeBot

Mobileye Global Inc.

Autonomous driving, ADAS chips, Level 4, EyeQ

Founded 1999 in Jerusalem; Intel acquired 2017 ($15.3B), returned public Oct 2022 (MBLY). Modular platform from driver assistance (SuperVision Level 3) to fully autonomous (Drive, Level 4).

Dominant ADAS chipmaker (EyeQ in millions of vehicles); scaling Level 4 with VW, Audi, Porsche and Uber, Lyft robotaxis.
Drive Level 4 testing in LA with Uber (2026); driverless service 2027. European trials (Oslo, Hamburg) starting 2026.
SuperVision (Level 3) and Chauffeur (Level 4 prototype with EyeQ6) in development; SOP 2027–2028 for premium brands.

mobileye.com (opens mobileye.com)Company profile →

Founded: 199927 yrs
Headquarters: Israel
Status: Public
Listed: MBLY (opens google.com)NASDAQ
Valuation: $7.1Bmarket capJul 2026
Last round: IPO (post-Intel separation) · $861MOct 2022
Works in: VehiclesRobotaxisAI
Coverage: 3 programs · 7 updateslatest 1 Jul 2026checked 10 Oct
Partners: Volkswagen Group (opens volkswagen.com)Uber (opens uber.com)Lyft (opens lyft.com)
Investors: Intel (opens intel.com)
People: Amnon Shashua (opens en.wikipedia.org)Founder; stepped down as CEO July 2026

Spec

MenteeBot

Training approach

Simulation + human demonstration generalization^R (reported)

Operation mode

Autonomous, no teleoperation required^R (reported)