Contact
CASE
Use Cases
What Is an AI Robot?
AI×ROBOT
What Is an AI Robot?
Industrial Robots Are Entering the Era of Thinking and Acting Autonomously
Recently, terms such as "Physical AI," "AI Robotics," and "Humanoid Robots" have become increasingly common in news reports and on social media. Industrial robots have traditionally excelled at repetitive tasks accurately and consistently as instructed by humans. In addition, Yaskawa has developed an industrial robot that can understand situations, make decisions, and act on its own through AI. This is the AI robot, "MOTOMAN NEXT."
Why are AI robots needed?
Because many tasks still require human skills.
Industrial robots have expanded mainly in manufacturing environments, including automotive plants, where they repeatedly perform the same tasks with speed and accuracy, such as welding and painting vehicles. This approach is called the "Teaching Playback Method."
The robots reproduce motions that humans teach in advance. Since then, industrial robots have been used in many more applications, including assembling home appliances and electronic components and sorting products in logistics warehouses.
However, many tasks still remain difficult to automate. Typical example is the task that require ambiguous or qualitative judgment.
For example:
- Handling food products with varying shapes and sizes
- Handling deformable objects such as cables and bags
- Tasks requiring human sensory judgment, such as determining whether something looks "good" or "acceptable"
These tasks have traditionally relied on human vision and intuition, making automation difficult.
Many activities that support our daily lives still depend on manual work, such as arranging prepared foods, routing cables, and packing strawberries neatly.
The AI robot "MOTOMAN NEXT" is designed to solve these challenges.
What can MOTOMAN NEXT do?
It can make decisions and act independently like a human.
Using information obtained from cameras and other sensors, MOTOMAN NEXT recognizes its surroundings, determines the current situation, decides what to do next, and plans and executes operations autonomously.
As a result, it can automate tasks where the position and shape of objects changes every time. This capability is known as robot autonomy.
The intelligence behind this autonomous recognition and decision-making capability is AI (Artificial Intelligence).
A Robot with Both Decision-Making and Execution Capabilities
Planning
The robot autonomously determines:
What is happening now? What should it do next?
It then generates an action plan.
Execution
MOTOMAN NEXT precisely performs the planned operation.
By adding AI-based object recognition and situational judgment capabilities, robots can perform tasks such as:
CASE
1
Identifying strawberries with different sizes and shapes and packing only those with desirable appearance.
CASE
2
Recognizing deformed cardboard boxes caused by transportation and opening them appropriately according to their condition.
CASE
3
Evaluating screw-tightening conditions, determining success or failure, and automatically retrying the task if necessary.
When AI performs the cycle of "Recognize → Decide → Plan → Act" in the physical world, it is called Physical AI. Within Physical AI, AI robotics focuses on robots. The technology that enables robots to autonomously "Recognize → Decide → Plan → Act" is called AI Robotics. Yaskawa utilizes Physical AI centered on AI Robotics and is advancing the implementation of MOTOMAN NEXT to automate tasks that have traditionally depended on human labor.
