Ascent Robotics
About
About Ascent Robotics
Ascent Robotics Corporation was incorporated in Tokyo in September 2016 and develops AI software for robots rather than a proprietary general-purpose hardware platform. Ken Kutaragi became chief executive in 2020, steering the company toward intelligent picking, simulation and digital-twin applications in logistics, manufacturing and pharmaceutical workflows.
AscentPick combines object recognition, grasp-pose inference, collision checking and motion planning in a web-configurable system compatible with multiple cameras, robot arms and hands. The company uses simulation to generate training data without requiring every object and scene to be captured and labelled physically. Partnerships with Bridgestone and Alfresa extend this stack into soft robotic manipulation and pharmaceutical logistics. Those collaborations support the commercial relevance of the software layer while keeping the company’s classification distinct from the robot and gripper manufacturers it integrates.
Supply
Supply chain relationships
- Zivid
Inbound - Sensing - Ascent Robotics integrated a Zivid 2 camera into its Ascent Pick robot-mounted picking application
Source
Team
- Ken Kutaragi - President and Chief Executive Officer
Timeline
Events
- 2025-03-04 - Partnership - Ascent Robotics and Alfresa sign pharmaceutical AI partnership
The programme combines Ascent's recognition and robotics software with Alfresa's pharmaceutical distribution network. Initial work targets digitalisation and automation in pharmaceutical logistics. The parties also identified future applications across medical, pharmacy and care settings.
Source - 2021-11-02 - Capability - Ascent Robotics integrates Zivid 2 into Ascent Pick
Ascent Pick uses the Zivid 2 for robot-mounted 3D perception. The system targets piece picking in manufacturing and logistics. The camera supports perception of small, shiny and reflective objects.
Source - 2020-12-16 - Product Launch - Ascent Robotics launches AscentPick robot-picking software
The system combines machine-learning recognition with a web-based configuration interface. AscentPick is designed to work with multiple robot, camera and gripper manufacturers. Simulation-generated training data reduces reliance on physical object capture and manual labelling.
Source
Context
Shared workflows
- Covariant
workflow adjacency - Warehouse AI Piece Picking
- Nomagic
workflow adjacency - Warehouse AI Piece Picking
- The Human Touch Robotics
workflow adjacency - Warehouse AI Piece Picking
- Sereact
workflow adjacency - Warehouse AI Piece Picking
- Dexterity
workflow adjacency - Warehouse AI Piece Picking
- CMES Robotics
workflow adjacency - Warehouse AI Piece Picking
Context tags
- Adaptive grasping - Capability - Verified against the company’s official product, technology and company materials during full-profile enrichment.
- Warehouse AI piece picking - Workflow - Verified against the company’s official product, technology and company materials during full-profile enrichment.