Mar 12, 2026 16:41 JST

Source: NEC Corporation

NEC Develops Physical AI That Anticipates Human Movement and Psychological States
Reducing Human Stress and Contributing to Human-Robot Collaboration

TOKYO, Mar 12, 2026 - (JCN Newswire) - NEC Corporation (NEC; TSE: 6701) has developed a first-of-its-kind physical AI (*1) that proactively controls robots to help reduce human stress. This technology utilizes a proprietary "world model" (*2) from NEC to predict human movement and psychological states, quantitatively estimating a person's level of stress.

The technology was developed based on a world model, an AI system that is attracting attention globally for accelerating robot adoption. It anticipates the direction a person will move based on the relative position and posture between a robot and the person, predicting the person's level of stress in real time. This enables robots to be controlled along paths and at speeds that reduce human stress.

In collaborative human-robot work environments, this reduction of psychological stress can be closely linked to physical safety, helping to avoid collisions and unexpected contact. As a result, expectations are growing for robots utilizing physical AI as a solution to labor shortages and hazardous tasks.

However, it is not easy for robots to control their movements based on expected human movements and psychological states. Moreover, robot behavior may cause humans to feel uneasy or tense, potentially hindering smooth collaboration.

NEC developed this technology to address these challenges. The technology promotes robot deployment in environments where dedicated robot zones are not yet established, such as small-to-medium-sized logistics warehouses, factories, and retail stores with narrow aisles—locations where stress-related barriers to robot use are conventionally high.

This development was achieved by integrating the knowledge and expertise NEC previously accumulated through the research and development of globally recognized models in the field of Physical AI and robot control technologies.

The proprietary world model utilized by this technology was developed by combining two prediction models. This has enabled robots to predict the future level of human stress when executing a specific driving control, thereby achieving optimal driving control that anticipates human movement. As a result, it can navigate at speeds and along paths that minimize human stress while maintaining mobility efficiency and suppressing unnecessary deceleration or stops, which can contribute to resolving labor shortages and improving productivity.

1. Predicting human movement that changes based on robot behavior and physical surroundings

By understanding the relationship between "human and robot behavior" as well as "humans and their physical surroundings," NEC has developed a unique prediction model capable of forecasting changing human movement. Using camera footage from robots and their control data, the model can predict with high accuracy the future 3D position and posture of people visible in current footage. This is accomplished in consideration of a robot's behavior and surrounding environmental information.

2. Real-time, quantitative estimation of the degree of human stress in relation to a robot’s position

Through the operation of robots and training AI with experimental results from stress surveys and robot movement data, NEC developed a unique predictive model that quantitatively estimates human stress levels based on the position, posture, and speed among humans and robots. This model enables real-time estimation of stress levels felt by individuals as robots approach them, tailored to each specific situation.

Going forward, NEC will continue contributing to the creation of environments where humans and robots can collaborate safely through the implementation of Physical AI.

(1) The amount of carbon dioxide absorbed and emitted between forests and the atmosphere.

(2) Source: Verra, a nonprofit organization that operates standards in environmental and social markets. "VCS VM0012 Improved Forest Management in Temperate and Boreal Forests (LtPF)."

About NEC

The NEC Group leverages technology to create social value and promote a more sustainable world where everyone has the chance to reach their full potential. NEC Corporation was established in 1899. Today, the NEC Group’s approximately 110,000 employees utilize world-leading AI, security, and communications technologies to solve the most pressing needs of customers and society. For more information, please visit https://www.nec.com, and follow us on LinkedIn and YouTube.

Source: NEC Corporation
Sectors: Artificial Intel [AI]

Copyright ©2026 JCN Newswire. All rights reserved. A division of Japan Corporate News Network.

Related Press Release


NEC and The University of Tokyo calculate carbon accumulation rates in Malaysian national park with a 9.4% margin of error
March 12 2026 11:02 JST
 
DOCOMO and NEC Launch Japan's First Commercial 5G Core on AWS, Built with World's First AI-automated Network Construction Technology
March 02 2026 15:34 JST
 
NEC Strengthens Multi-Vendor Optical Network Solutions with New TIP-Certified Phoenix Hardware Lineup
February 27 2026 17:50 JST
 
NEC Demonstrates Agentic AI-Driven Autonomous Network Operations in Collaboration with AWS
February 27 2026 16:31 JST
 
NEC Strengthens Its Edge Portfolio for the 6G Era, Enabling Efficient Edge Deployment and Integration with Cloud Services
February 27 2026 11:24 JST
 
The University of Tokyo, NTT, and NEC demonstrate real-time augmented reality assistance made possible by integrating three newly proposed technologies on a 6G/IOWN platform to realize the widespread use of AI agents supporting safety and security
February 26 2026 18:01 JST
 
NEC and Nokia to Expand Eletronet's Optical Fiber Network in Brazil by 50%
February 26 2026 17:51 JST
 
NEC Implements AI Code Review Service "Metabob," Reducing Technical Verification Time by Up to 66%
February 25 2026 17:00 JST
 
NEC Exhibits at MWC Barcelona 2026, the World's Largest Mobile-Related Exhibition
February 09 2026 20:18 JST
 
NEC Develops High-Efficiency, Compact Power Amplifier Module for 5G Base Station Radio Units
January 28 2026 09:05 JST
 
More Press release >>

Latest Press Release


More Latest Release >>