NXP chief says robots need autonomous computing to move beyond remote control

NXP’s CEO has exposed the core bottleneck in today’s robotics: dependence on remote control masks a lack of true autonomy. Technically, this shift demands co-evolution of edge AI accelerators, ultra-low-power heterogeneous SoCs, and deterministic real-time OS—favoring vendors with integrated safety islands. Regulatory pressures from the EU AI Act and U.S. export controls are escalating compliance costs, particularly around algorithmic transparency and data sovereignty; foundries in Taiwan, China lacking ISO/SAE 21434 certification risk exclusion from high-end robotic BOMs. Competitively, NVIDIA dominates perception via Orin, while Renesas and STMicro accelerate vertical integration of motor-control-to-power ICs to bypass NXP’s stronghold in industrial communication protocols. Within 18 months, only chips enabling sub-watt on-device inference—and eventually on-edge fine-tuning—will define commercial viability; others will remain stuck as demo curiosities.