As automation accelerates across global industries, the physical safety of human workers has always been the top priority. However, the psychological impact of working alongside large, autonomous machines is often overlooked. Addressing this critical gap, NEC Corporation has developed a first-of-its-kind physical AI designed to proactively control robots with the specific goal of reducing human stress.
At AarokaTech, we closely monitor the evolution of industrial automation. This breakthrough from NEC represents a massive shift in human-robot collaboration. By utilizing a proprietary “world model,” these new robots can predict human movement and quantitatively estimate a person’s psychological state in real time.
The Psychological Barrier to Robot Adoption
While robots are highly effective solutions for labor shortages and hazardous tasks, integrating them into human-centric environments is complex. Even if a robot is programmed to avoid physical collisions, its sudden movements, speed, or proximity can cause humans to feel tense, uneasy, or distracted. This psychological stress acts as a major barrier to smooth collaboration and widespread deployment.
NEC developed its new physical AI to dismantle these barriers. This technology is particularly highly anticipated for environments where dedicated, caged-off “robot zones” are impossible to establish. Examples include small-to-medium-sized logistics warehouses, bustling factory floors, and retail stores with narrow aisles.
How the NEC “World Model” Works
To solve the anxiety issue, NEC integrated years of accumulated expertise in physical AI and advanced robot control technologies. The result is a proprietary “world model”—an AI architecture that is rapidly gaining global attention for its ability to accelerate robot adoption.
The NEC world model is built upon two distinct, highly sophisticated predictive engines:
1. Predicting Dynamic Human Movement
By analyzing the complex relationship between human behavior, robot behavior, and the surrounding physical environment, NEC has developed an AI capable of forecasting how a person will move.
Using live camera footage fed directly from the robot, combined with internal control data, the model accurately predicts the future 3D position and posture of the people in its vicinity. It doesn’t just see where a human is; it anticipates where they are going and how they will react to the robot’s presence.
2. Real-Time, Quantitative Stress Estimation
Predicting movement is only half the solution. NEC took this a step further by training their AI with experimental data from comprehensive stress surveys and corresponding robot movement logs.
The resulting predictive model can quantitatively estimate human stress levels in real time. It calculates this based on the relative position, posture, and approach speed between the human and the robot. As the machine approaches a worker, the AI instantly gauges the potential stress level tailored to that exact, split-second situation.
Optimizing Productivity Through Empathy
By combining these two models, NEC’s physical AI robots can navigate facilities along paths and at speeds that actively minimize human anxiety.
Crucially, the system achieves this psychological safety without sacrificing operational efficiency. Because the AI accurately anticipates human movement, the robot can avoid unnecessary, jarring decelerations or abrupt stops. It maintains a smooth, predictable workflow that resolves labor shortages and improves overall productivity while keeping the human workforce comfortable.
Going forward, NEC plans to continue refining its physical AI technologies, contributing to a future where humans and robots can collaborate not just safely, but seamlessly.
