Google DeepMind Accelerator: Robotics Program Launch

Google DeepMind Accelerator: Robotics Program Launch

Google DeepMind has launched the Google DeepMind Accelerator: Robotics, a three-month program designed to provide early-stage robotics startups across Europe with technical mentorship, product guidance, and access to Google's AI stack and Gemini robotics models. This initiative aims to transition cutting-edge AI research into real-world applications across diverse sectors including healthcare, logistics, and manufacturing.

Program Structure and Support

The Google DeepMind Accelerator: Robotics provides selected startups with intensive support over a three-month period. The program focuses on enabling the integration of AI into core products through the following resources:

  • Technical Expertise: Hands-on support and mentorship from Google and Google DeepMind experts.
  • AI Infrastructure: Access to the Google AI stack and specialized Gemini robotics models.
  • Product Guidance: Strategic guidance to help founders scale their operations responsibly and optimize their product development.

Participating Robotics Startups

The inaugural cohort consists of 15 companies from across Europe, focusing on a wide range of embodied AI applications:

Industrial and Manufacturing

  • 3D-Components AS (Norway): Automates parameter selection and quality control for robotic welding and metal 3D-printing via the RobTrack platform, claiming speeds 280X faster than current practices.
  • Acumino (Greece): Develops hardware-agnostic Physical AI for scalable and cost-efficient industrial tasks.
  • Deltia GmbH (Germany): Digitizes production-line work by transforming workflows into process graphs to optimize manual processes.
  • AUAR (United Kingdom): Deploys robotic MicroFactories to construction sites to make homebuilding more affordable.

Healthcare and Biotechnology

  • Adapta Robotics (Romania): Uses physical AI to replicate human touch for testing devices and software in the automotive, consumer electronics, and healthcare sectors.
  • ROBEAUTE (France): Develops microrobots for navigating brain tissue to diagnose, treat, and monitor neuropathology in neurosurgery.
  • Touchlab (United Kingdom): Creates high-resolution "e-skin" using advanced nano inks to provide robots with a sense of touch.

Logistics, Environment, and Navigation

  • Bubble Robotics (France): Creates a vessel-free constellation of self-docking surface and subsea robots to build a live underwater world model.
  • Danu Robotics (United Kingdom): Automates complex waste sorting using embodied AI to improve material recovery and support the circular economy.
  • Staer (Sweden): Uses computer vision to build 3D spatial models of facilities for robot navigation and operational visibility.

General Purpose and Foundation Models

  • Embodied AI (Switzerland): Employs teleoperated humanoids to collect data for training manipulation skills.
  • Extend Robotics (United Kingdom): Provides teleoperation software and data pipelines to fine-tune foundation models for robotics.
  • Forgis (Switzerland): Develops AI agents that predict machine failures and optimize operations.
  • Generative Bionics (Italy): Develops humanoid robots based on physical AI.
  • Qualia (Denmark): Builds infrastructure to convert robotic foundation models into working deployments.

Sources