Sat Nov 08 05:20:00 UTC 2025: News Article:
Experts Urge Caution as Physical AI Robots and Drones Become More Prevalent
The rise of robots and drones powered by physical AI is rapidly transforming industries, from logistics to healthcare. While humanoid robots with advanced artificial intelligence are still on the horizon, application-specific robotics are already becoming commonplace. However, experts are warning that extra safety measures and careful consideration are needed to prevent accidents and bias as these systems become more integrated into daily life.
“It’s too early for us to know how robot and human interactions will play out,” said Sathishkumar Balasubramanian, head of product for IC verification and EDA AI at Siemens EDA. “People are very cautious in how they do it, how they interact, because it all comes from understanding the automated robot perspective of the object it is interacting with.”
The challenges stem from the complex interplay of hardware, software, and AI models that power these systems. Advanced sensors, AI cameras, and large language models (LLMs) enable robots to interpret their surroundings and make decisions, but potential failures and biases in these components could lead to unintended consequences.
“Rigorous testing needs to be performed not just on the hardware or other processors but also on the AI,” said Rajesh Velegalati, principal security analyst at Keysight Technologies. “I often observe that for the same prompt, I don’t get exactly the same output… this should not be the case in safety-critical applications.”
To address these concerns, researchers are focusing on developing robust verification techniques, AI auditing tools, and simulation environments. These tools enable engineers to identify potential flaws and biases in physical AI systems before they are deployed in real-world scenarios.
“We need to think about competency management and focus on people, processes, and development frameworks, not just the tools and technology,” said Andrew Johnson, engineering and technology leader, systems and functional safety engineering specialist at Imagination Technologies.
Simulation is also playing a crucial role in training AI models for robots and drones. By creating physically accurate virtual environments, engineers can test and refine these systems under a wide range of conditions.
Experts agree that the physical AI sector is still in its early stages of development, but the potential for innovation is immense. As robots and drones become more sophisticated, ongoing research and collaboration between industry and academia will be essential to ensure their safe and responsible integration into society.
First Piper 