What Do Star Wars, Raspberry Pis, and Your Mom Have in Common? The Distributed Revolution in Mobile Robotics

I'm back with a new season of Conversations at the Edge (CATE), and this time, we're diving into the fast-moving world of mobile robotics. Last season, we stirred the pot with Matter in our smart homes, but now we're going deeper into the machines that are rolling into warehouses, hospitals, farms, and maybe even your local delivery service.

Our first mobile robotics topic for discussion? The Distributed Revolution in Mobile Robotics. If that sounds grand, it's because it is. Centralized "brains" are giving way to distributed intelligence where sensors, processors, and controllers all share the load. For developers, this shift changes how you design architectures, plan for safety, and think about scaling your platform.

And I'm not doing it alone.

Meet my new co-hosts

To keep things fresh, I've got two brilliant co-hosts on board.

Together, Altaf and Iain keep me honest, challenge assumptions, and bring decades of insight into what it really takes to build robots that move, think, and survive in the wild.

Episode 1: MVPs are a Trap

We poked the bear with this one. Minimum Viable Products (MVPs) sound great, but, in robotics, they can trap you. If you design for "just enough to demo," you'll hit a wall when it's time to scale.

Instead, think platforms. Modularity and scalability let you add new sensors, swap in higher-performance compute, and layer on revenue-generating features without starting over. We showcased our B3RB buggy reference design, which runs ROS 2 and demonstrates how distributed compute between a main processor and companion microcontrollers creates a scalable platform. The point for developers: don't build one-off products, build architectures that can evolve for years.

Watch the Episode Now

Episode 2: You Can't Go to Market with a Raspberry Pi

I love Raspberry Pi for tinkering but it's not the path to production robotics. Industrial robots need long-term supply, safety certifications, and hardened security. Prototyping boards for consumer applications can't provide that.

Developers need silicon that's designed for real environments. For example, i.MX RT crossover MCUs for drones, automotive-grade processors with lockstep cores and secure elements, and full reference designs that integrate motor control, battery management, and connectivity. These give you a launchpad to market-ready systems without the headaches of trying to scale hobby boards.

Watch the Episode

Episode 3: Mobile Robotics, AI, and Your Mom

Yes, we really went there. Iain tied a lesson from his mom, "be home by dark", to AI. The analogy? Let your AI explore but always set boundaries.

The three trends we called out in this episode are shaping robotics today:

• Agentic AI: robots that don't just perceive, but act. For example, detecting a fire and triggering a response
• Hardware safety: processors supervising processors, cutting power to motors if limits are exceeded
• Edge intelligence: neural accelerators like those in i.MX 95 pushing decisions closer to the sensors for faster, more autonomous responses

For developers, that translates into designing systems where intelligence is layered, supervised, and distributed across the architecture. And if you still think AI is "just software," our NavQPlus reference design, running Linux, vision pipelines, and dual neural accelerators, proves otherwise.

Watch the Episode

Episode 4: Star Wars Got It Right with Distributed Architecture

Yes, we compared today's mobile robotics to Star Wars. Just look at the movie's R2-D2, BB-8, or even those mouse droids scurrying around: each has one job that they do really well. That's the essence of distributed robotics and this episode.

For developers, the lesson is to architect platforms that scale across robot types. A humanoid assistant, a four-wheeled warehouse rover, and a quadcopter drone may look different, but they all need distributed decision-making. NXP solutions like the i.MX 8M Plus (for vision and neural acceleration), the S32K1 (for real-time motor control), and Automotive Ethernet (for lightweight, deterministic networking) provide the building blocks. The challenge is assembling them in the right distributed mix.

Watch the Episode

Episode 5: A Centralized Architecture is Dangerous

We opened this episode with a warning: robots aren't caged or limited to a specific area anymore. They're navigating warehouses, rolling through loading docks, and interacting with humans. A centralized architecture simply can't equip robots with what they need to keep up with unpredictable hazards.

The main takeaway? Safety has to be distributed. That means embedding perception directly into sensors (vision, LiDAR, radar) and backing it with supervisory processors that can overrule the main brain if it makes a bad call. Think of it like a reflex: your hand pulls away from a hot stove before your brain even processes it. For developers, that means working with platforms like NXP's edge processors that enable this kind of local intelligence and fail-safe behavior.

Watch the Episode

What it means for developers

Across these first episodes, one theme stands out: distributed intelligence isn't a buzzword. It's a critical strategy.

Robots are leaving the lab and entering messy, unpredictable environments. Developers need to build with architectures that perceive, decide, and act locally, without waiting for a central brain to figure it out. That requires modular platforms, scalable hardware, and robust safety frameworks.

Which is exactly what we explore in NXP's new white paper, From Stationary Arms to Humanoids: The Journey of Mobile Robotics. It examines control complexity, AI adoption, and modular robot architectures to offer distributed intelligence as the path forward.

What's next on CATE

This is just the first segment of our Mobile Robotics discussion. Each month, we'll release a new set of episodes tackling another facet of robotics. So, if you liked the mix of bold statements and grounded engineering insights, stay tuned for more and let us know what you think.

And seriously, don't just take my word for it. Head over to YouTube to watch the full conversations, see our demos, and maybe even laugh at our analogies (yes, your mom really does belong in this discussion).