Intelligence,into the physical world.
Physical AI that adapts to the messiness of the real world — industrial robots made autonomous by VLA foundation models.
What is Physical AI
Physical AI is technology in which digital intelligence acts on the physical world through embodied systems such as robots. On factory floors it enables flexible responses to irregular situations and automates expert skills — improving productivity and easing labor shortages.
Why Cistron
Research-Backed
Direct line to academia
Our CTO is an Assistant Professor (Ph.D.) at Kanazawa University. State-of-the-art robotics theory goes straight from the lab into deployment — no translation layer.
End-to-End
From data to deployment, one team
Acquisition, Adaptation and Actuation handled by a single company — no precision lost in the gaps between vendors.
Field-Ready
Built to enter real factories
With our production partner Mikawa Sangyo, we cover equipment design and build, electrical and piping work, and commissioning.
Foundation Models
Adaptive intelligence via VLA
VLA foundation models built on imitation and reinforcement learning adapt to the irregularity of real production floors.
From challenge to factory adoption
Understand & design
We study the target process and its safety constraints, then design the sensing and data collection needed for the task.
Train & validate
We adapt models with real data and validate accuracy, speed, and safety on physical equipment before production use.
Deploy & improve
We integrate with existing equipment, deploy to production, and continuously improve performance using operational data.
Pricing
R&D
by Cistron Inc.
- Problem discovery & requirements
- Core technology R&D
- Software / hardware design
- Proof of concept (PoC)
- Continuous PDCA iteration
System Install
by Mikawa Sangyo Co., Ltd.
- Production equipment design & build
- Site preparation
- Removal & installation
- Electrical & piping work
- Commissioning & tuning
Academic insight, deployed to the real world.
Led by CTO Dr. Nishimura (Assistant Professor, Kanazawa University), we fuse state-of-the-art robotics theory from the university lab with deep expertise in real-world physical constraints — delivering precise, robust Physical AI backed by academic rigor.
Naruhiro Nishimura
CTO / Assistant Professor, Kanazawa University · Ph.D.
Join us
Physical AI Engineer
We are hiring engineers to deploy Physical AI into major automotive plants. Apply by email with your profile.
Latest updates
Cistron Accepted into the Onshape Startup Program→
CistronがOnshape Startup Programに採択されました。クラウド型3D CADと設計データ管理を活用し、ロボット開発における設計・試作・チームでのレビューを支える環境づくりを進めます。
Sim2Real — Building an Isaac Simulator Environment on AWS→
NVIDIA Omniverse 基盤のシミュレーション環境を AWS 上に構築するサービスについて紹介します。
RTJ2026: AI Robot Collaboration with YUASA and NIT Toyota College→
RTJ2026の共同展示に向け、Cistronと豊田高専の学生が音声でノベルティを届けるAIロボットを開発。共同開発の役割分担、展示概要、日刊工業新聞などの関連リンクを紹介します。
Before we talk.
Can we talk before requirements are defined?
Yes. Our work includes problem discovery and requirements definition. Tell us about the current task and why you want to automate it.
What should we share?
Describe the task, objects, existing equipment, and current difficulties as far as you can. Share only information you are able to disclose.
How much does development cost and how long does it take?
This depends on the data needed, physical testing, and equipment integration. We define the development scope with you and provide an individual quote.
Tell us what is difficult
on your factory floor.
“Could this task be automated?” is enough to start. Share your process and the challenges you are facing.