NXP Semiconductors has launched a unified robotics platform that bundles chips, an open-source software kit and ready-made drone designs into one package. The Dutch chipmaker announced the platform on 9 November 2026, and it is available now.
The platform is built on what NXP calls its Neural Axis Architecture, a distributed foundation that links sensing, real-time control, AI processing and communications across a robotic system. The idea is to give developers one validated route from a single function to a complete machine.
A drone in 30 minutes, out of the box
The headline claim is speed. NXP says developers using its reference designs can deliver initial drone functionality within 30 minutes out of the box. The company frames that against the months teams usually spend integrating and validating robotic subsystems.
The first reference designs cover four drone subsystems. Autonomous compute and flight control run on the NavQ95, built around the i.MX 95 applications processor. Flight management uses the Tropic VMU and FMU, based on the i.MX RT1064 crossover microcontroller.
Sensor networking runs on a 100Base-T1 sensor node built on the MCX N947 microcontroller, while motor control uses an AM32-compliant electronic speed controller based on the MCX A13 microcontroller. Each design carries the safety and security features of its main processor.
An open-source SDK built on familiar frameworks
The software side is a modular, open-source kit with drivers, middleware integration, reference applications, configuration tools and reusable components. It works with PX4, Zephyr and ROS 2, three frameworks robotics developers already use.
The kit also bundles ROS 2 Jazzy, the PX4 autopilot, CogniPilot Cerebri, Zephyr RTOS and NuttX, plus the eIQ Toolkit and Ara SDK for edge AI work. Developers build on a Linux host with Docker, using the west meta-tool and the Arm GNU toolchain.
Drones first, service robots later
For now the platform speaks to drone builders only. Service robots, autonomous mobile robots and other autonomous systems get their own reference designs in the second half of 2027, according to the announcement.
That gap matters for startups eyeing warehouse or hospitality robots. The software foundation is ready today, but the validated designs for their machines are still more than a year away.
Two drone companies already back the approach. James McClearen, chief technology officer at blueflite, called the platform a ready-to-build foundation that shortens development cycles.
Srdjan Kovačević, chief executive and co-founder of Orqa, said working with NXP at the architectural level gives his team a stronger base for its next generation of systems. Developers can find the SDK and the drone reference designs on NXP’s robotics pages.
Why NXP is betting on physical AI now
NXP is selling the picks and shovels of the robotics boom rather than the robots themselves. The company posted revenue of 12.27 billion dollars in 2025 and operates in more than 30 countries, with its chips already inside cars, factories and connected devices.
The market it is chasing is growing fast. Global shipments of professional service robots rose 24 percent to almost 250,000 units in 2025, according to the International Federation of Robotics. Transport and logistics robots led the way with 117,500 units sold.
Consumer service robots grew even faster, up 37 percent to almost 34.2 million units in 2025. That demand is pulling chipmakers toward robotics, and NXP wants its hardware and software to be the default starting point.
KeeVurds has tracked the same shift from the startup side. Physicl raised 10 million dollars in November to train robots on the rare situations they keep getting wrong, a sign that the hard part of robotics is now software and edge cases rather than hardware alone.
Trust is the pitch
“As autonomous robots increasingly move beyond controlled environments and into the spaces where people live and work, trust becomes a prerequisite for adoption.”
Jerome Legros, Vice President and General Manager, Robotics, NXP Semiconductors
Legros argues that security, functional safety, real-time control and lifecycle reliability have to be designed in from the start. That is the thread running through the whole launch, and it is also the test the platform will be judged on as robots move closer to people.
