Digital Twin
Recreates real agricultural environments in high-fidelity virtual spaces where AI robots can learn and be validated. Fields, crops, equipment, and robot motion are digitized so hard-to-capture field data can become reusable assets.
Physical AI Development Platform
Accelerating Physical AI
with Digital Twins × World Models
By 2030, Japan’s agricultural workforce is expected to shrink dramatically.
The next step is a development platform for AI robots that can operate in the real world.
From agricultural simulators to a Physical AI platform
About Smart3tene
From AI that thinks to the development of AI robots that act
Smart3tene recreates agricultural environments in digital space to support AI training, validation, and deployment on real machines.
World Model
Enables AI to understand and generate changes in weather, crop growth, obstacles, and work conditions. It forms a decision-making foundation for robots that must respond to unfamiliar real-world situations.
Sim2Real
Connects AI models trained and validated in virtual environments to physical robots. This reduces development time and field validation cost while accelerating real-world deployment of AI robots.
Physical AI
Provides a foundation for AI robots to make autonomous decisions and act in the real world, based on intelligence gained from digital twins and world models. Starting with agriculture, it can expand to many industrial settings.
R&D and Results
Technical Development Through Joint Research
Smart3tene advances agricultural robotics, digital twins, and AI simulation technologies through joint research with universities and research institutes.
Joint Research
Collaborating with universities and research institutes to advance AI robot technologies for agricultural sites.
Digital Twin Development
Recreating fields, crops, and terrain in virtual space to build AI learning environments.
Robot Applications
Researching harvesting, recognition, mobility, and work support with real-machine deployment in mind.
Joint research with Akita Prefectural University
Virtual Tomato Farm Simulator
Developed a simulator that recreates a tomato farm in virtual space, including automatic annotation, MR harvesting experiences, VR robotic harvesting, and automated harvest simulation.
Joint research with Akita Prefectural University
Remote Mower Training Simulator
Developed a web-based simulator for remote mower operation training, along with a remote mowing website and an operating system for remote mower control.
Joint research with Akita Prefectural University and JA Aso
Strawberry Sorting Robot Simulator System
Developed a virtual simulator for a strawberry sorting robot that uses AI and sensors to recognize individual berries and support robot applications for gentle sorting and packing.
Digital Twin Environment for Agricultural Robots
Developing a 3D simulation space for reinforcement learning in robot control and for validating robot work in agricultural environments.
Construction Support Technology Based on Microtopography
Joint development has begun on an Android smart drainage guidance app, with field validation and an Android app release planned for fiscal year 2028.
Use Cases
Physical AI Applications Starting in Agriculture
Smart3tene technology can be applied beyond agricultural robots to many fields that need AI systems operating in the real world.
Agriculture
Advancing smart agriculture
and agricultural robots
World Models & Physical AI
Advancing smart agriculture
and agricultural robots
Logistics, Construction & Healthcare
Expanding into a wide range
of real-world industries
Move Real-World AI Development to the Next Stage
For technical consultation, joint research, or business partnerships related to Smart3tene,
please contact us.