"The semiconductor technology that has led innovation in the auto industry over the past 10 years will accelerate the future of Robotics."
Park Jung-seo, head of Texas Instruments Incorporated (TI) Korea, said this in a keynote speech at the "TI Mobility & Robotics Seminar 2026" held at EL Tower in Yangjae-dong, Seoul, on the 9th. He said the semiconductor platforms proven in electric vehicles and self-driving cars can speed up the commercialization of various robots, including humanoids. Park said, "Cars and robots are increasingly sharing similar system-level challenges," adding, "TI is helping engineers develop next-generation intelligent machines based on proven analog and embedded processing technologies."
TI Korea held technical presentations and demo booths that day on the theme of convergence between the auto and Robotics industries. It introduced how automotive semiconductor technologies—such as real-time actuation loops (a control flow that continually reflects sensor information in motor movement), motor control, sensing, communication, radar, vision, Edge AI (AI that processes information on the device to perform actions), and 48-volt (V) power systems—apply to robot design. About 200 industry officials and researchers attended.
Founded in 1930, TI is a U.S. company focused on analog semiconductors and embedded processing chips. Last year's revenue was $17.682 billion (about 26.615 trillion won), and operating profit was $6.023 billion (about 9.069 trillion won). TI has been in the automotive semiconductor business for more than 40 years.
◇ TI "Humanoid market at $51 billion in 2035… Korea's production share seen at 30%"
Market research firm Yole Group projected that the global Humanoid Robot market will grow from $600 million (about 903 billion won) in 2025 to $51 billion (about 76.765 trillion won) in 2035. TI forecast that Korea, with its base in autos, electronics, batteries, manufacturing, and industrial automation, could account for 30% of global humanoid production in 2035.
TI's strategy is to deliver results in the growing humanoid market with automotive semiconductor technology. With the advent of electric vehicles and software-defined vehicles (SDVs), the share of semiconductors installed in batteries, motors, sensors, communications, and power control has increased significantly. Humanoids similarly require sensor, joint, battery, and safety control circuits, making them technically akin to cars.
Park said, "The critical elements in the evolution of robots—perception, control, safety, and high-efficiency power—are technologies that have already advanced in cars over the past several decades," adding, "The trajectory in which electric vehicles have evolved with charging infrastructure, batteries, 48V power architectures, Autonomous Driving, and vehicle-to-everything (V2X) can apply to Robotics."
Park saw four challenges that must be solved for humanoids to spread: perception, manipulation, safety, and power. Park said, "People easily pick up objects, but it isn't easy for a robot to naturally perform actions like moving an egg or throwing a ball," adding, "When humanoids assist people, safety cannot be overemphasized."
◇ "Car semiconductors that move robot joints… real-time control is key"
Heo Jeong-hyeok, a director at TI Korea, gave a presentation on "Ethernet-based real-time actuation loops for next-generation mobility and Robotics." He said, "When you combine electric vehicle motor traction with precision control solutions for industrial automation, you get Robotics," adding, "Even within TI, it's not that we prepared separately for Robotics; our capabilities in automotive semiconductor technology were already strong enough to say we were quite well prepared."
In humanoids, actuation—the process of converting electrical signals into actual movement—is considered a core technology. A humanoid's 48V battery ties into vehicle power systems, the battery management system (BMS), and power distribution technologies. Motor control for moving joints is structurally similar to the traction inverter in electric vehicles.
Heo said, "To precisely control a motor to the desired position, the control speed must first be fast," adding, "TI microcontroller units (MCUs) can perform the FOC computation, commonly used in motor control, in about 500 nanoseconds (ns), and high-end products can process everything from sensing to computation to reflecting PWM in no more than 1 microsecond (μs)." FOC is a method that controls the direction of the motor's magnetic field and current to improve efficiency and precision.
Ethernet-based communications, an internal network where controllers, sensors, and motors—mainly used in Autonomous Driving cars—exchange data rapidly, is also being applied to humanoids. Heo said, "The most watched next-generation communication in robots is EtherCAT (a real-time Ethernet communication method widely used in industrial automation)," adding, "Each module and inverter can exchange data with high bandwidth and move smoothly with low latency."
◇ Demonstrations of semiconductor technologies applicable to self-driving cars and humanoids
The technology demo booths featured ▲ semiconductor boards for robot joint control ▲ 48V battery management systems ▲ radar and vision sensor fusion ▲ SDV zonal architecture demonstration equipment, among others. The robot joint control demo introduced gallium nitride (GaN)-based power devices. GaN is a material advantageous for high-speed switching and high-efficiency operation compared with conventional silicon Power Semiconductor. A TI Korea official said, "Using GaN reduces heat generation and can shrink the board size by about 50% while driving the same power."
The 48V battery management system demo also drew attention. For robots to move, they need to carry a battery pack, and the voltage, temperature, and status of each cell must be managed in real time. A TI Korea official said, "Robots also have battery packs, and a chipset to manage lithium-ion batteries is mounted on the board," adding, "It communicates with the MCU to perform cell balancing or detect fire risks in advance."
In the automotive 48V zonal architecture demo, a wiper actuator was used. Conventional vehicles are organized around controllers by function. Zonal architecture divides the vehicle by physical location and connects nearby sensors and actuators to the corresponding controller. Park said, "When the structure managed by function changes to one grouped by location, efficiency improves for the wire harness."