Topic Questions
Robotics have changed in the past decade, older industrial robots needed to be physically fenced off to “Smarter Robots” (like Cobots and AMRs). What new design considerations come along with these smarter robots?
-
Traditional industrial robots have been used for decades and excel at repetitive, heavy, and hazardous tasks, but they require tightly controlled environments and are unsafe for humans to work near. As robots are increasingly expected to operate alongside people, safety becomes the primary concern. This shift is enabled by adding many more sensors—such as ultrasonic imaging, LiDAR, and radar—which are combined through sensor fusion so algorithms can accurately perceive and navigate dynamic environments.
-
Unlike earlier industrial robots that were largely stationary and task-specific, newer “smarter” robots are designed to move freely and adapt to changing surroundings. This flexibility allows them to handle a wider range of tasks beyond traditional applications like lifting heavy objects, welding, or painting in automotive manufacturing. With advanced sensing and physical AI, a single robot can now be repurposed throughout the day—for example, tending a CNC machine in the morning and handling packaging later. This adaptability lowers the barrier to adoption, especially for smaller companies, by reducing the need to redesign entire production lines and enabling incremental automation.
We have seen what Autonomous Mobile Robots (AMRs) can do, such as self-navigate and even relocate shelves in fulfillment centers. Where do you see AMRs going next?
-
Robotics adoption is starting in controlled environments like warehouses, where lighting and conditions are predictable and simpler imaging sensors can be used. As robots move into less controlled outdoor environments—such as agriculture or delivery—they require more advanced sensing, including high dynamic range imaging to handle variable lighting, as well as force feedback, rotational positioning, and moisture sensors to perform delicate tasks like picking fruit. This need to handle environmental variability is driving a rapid expansion in sensor technologies.
-
In the near term, robotics will continue to grow fastest in industrial settings, where robots can safely and efficiently perform many tasks currently done by humans. A later wave will bring robots into homes to handle physically demanding, unpleasant, or dangerous chores, particularly as populations age. Robots are also well suited for hazardous and security-related jobs.
-
Safety is a major driver of adoption, especially in applications like autonomous forklifts, where high accident rates and workforce turnover make automation attractive. Robots are often deployed where tasks are too dangerous, large-scale, or risky for humans. Technologically, mobile robots share many similarities with autonomous vehicles and ADAS systems, but operate at much lower speeds, reducing risk. In industrial settings, robots are designed to slow down or stop when encountering uncertainty, a safety-first approach that continues as robotics expands into new domains.
Navigation & autonomous driving is not a trivial challenge currently. I saw that you used SLAM (Simultaneous Localization & Mapping) to solve this problem. I know there’s a lot of hot topics that play into this (like Environment modeling, Obstacle detection, Dead Reckoning, and Optimal path planning). Can you tell me more in your own words how this all comes together?
-
Robotic perception and learning can be compared to a blindfolded person navigating a familiar versus an unfamiliar room. The major innovation enabling robots to handle unfamiliar spaces is the use of digital twins in simulation environments (such as an Omniverse), where robots can train virtually through massive trial and error. By simulating rooms, warehouses, or entire facilities, robots—or fleets of robots—can practice hundreds of thousands of workflows overnight, learn collaboratively, and avoid collisions before ever being deployed in the real world.
-
This virtual-first training gives robotics an advantage over automotive autonomy, where the “first vehicle through” a newly changed road must react in real time and then propagate updates to others. As a result, automotive autonomy is advancing fastest in controlled routes like freight and yard operations, similar to how industrial robotics benefits from controlled environments.
-
On the technical side, modern mobile robots rely on platforms like ROS combined with high-bandwidth sensor integration technologies such as NVIDIA Holoscan, which reduce latency by moving sensor data directly into memory for processing. Low latency and high bandwidth are critical not only for navigation but also for emerging applications like telemedicine and remote robotic surgery.
-
Robotic control architectures are increasingly split into “system one” and “system two”. System one handles low-level autonomy such as balance, posture, and basic motion, allowing robots to remain stable without human intervention. System two manages higher-level tasks like navigation and goal execution, enabling clear separation between reflexive control and intelligent decision-making.
-
Advances in sensing are also key, particularly in-depth perception. Traditional cameras provide only 2D images, so new technologies—such as indirect time-of-flight (iToF) cameras that emit safe lasers to measure depth—add critical spatial awareness. These sensors allow robots to understand distances and relationships between objects more accurately. Together, improvements in sensing, simulation, system architecture, and AI-driven learning are rapidly maturing robotics and accelerating their ability to operate safely and autonomously alongside humans.
YouTube Video Sources
If you would like to watch this video or any other Let’s Talk Technical Series, please utilize the sources below.
Video on this Topic
Let’s Talk Technical with onsemi: Robotics and Physical AI | DigiKey
Other Videos from this Series
Let’s Talk Technical Playlist - YouTube
Supplier Featured in Video
onsemi
onsemi is driving energy efficient innovations, empowering customers to reduce global energy use. The company offers a comprehensive portfolio of energy efficient power and signal management, logic, discrete and custom solutions to help design engineers solve their unique design challenges in automotive, communications, computing, consumer, industrial, LED lighting, medical, military/aerospace and power supply applications. onsemi operates a responsive, reliable, world-class supply chain and quality program, and a network of manufacturing facilities, sales offices and design centers in key markets throughout North America, Europe, and the Asia Pacific regions.