Can AI Design IoT Hardware? | Flux's Matthias Wagner | Internet of Things Podcast

IoT For All Podcast · 14 Apr 2026 · 21m · Episode 397

AI in practice: 3 real uses of AI described

What this episode covers

Matthias Wagner, Founder and CEO of Flux, explains how his company uses agentic AI to democratize PCB design, allowing non-engineers to create hardware via natural language prompts. He discusses the historical bottlenecks of hardware design, the necessity of iterative back-and-forth for complex devices like smartphones, and the economic advantages of designing custom boards for products like vending machines. Wagner also outlines future integration challenges, including firmware generation and enclosure design, and predicts that lowering the cost of iteration will unlock significant innovation in robotics and IoT.

AI in practice

Where this conversation describes AI being used to build or deliver a product. Part of the AI in practice tracker.

DesignFlux

The AI agent interacts with non-engineers to clarify design requirements (such as pipe material and sensor type) and then generates a complete PCB design for specific applications like farm automation.

“but once you ask them these questions, then flux can definitely design you the board and get that manufactural with you together. Right? But even for like a simple thing like that, there's a bunch of like detailed questions, right?”

Matthias Wagner · Hear it at 5:21

DesignFlux

The AI agent assists users by expanding trade-offs between different hardware options, such as power sources, explaining pros and cons to help users make informed design decisions.

“I mean, the agent can expand these trade-offs to you, right? The difference between USPC and USBA or, you know, or any other power source for that matter? What the pros and cons are and what the best fit is for your application, right?”

Matthias Wagner · Hear it at 9:07

DesignFlux

A vending machine manufacturer used Flux to design a single integrated PCB that replaced four or five separate boards, reducing assembly complexity and total cost of ownership.

“But that's kind of like the sales quo. And so then what we see is like, we have this customer, they switch to flux. They made themselves a single integrated port with flux. Right? So now also think about the assembly of the vending machine on the assembly line is one border put in.”

Matthias Wagner · Hear it at 9:07

Mentioned

Companies
Flux, Siemens, GitHub, Home Depot
Products
ChatGPT, Roomba
Chips
ESP32

Other shows on the same ground, closest first.

  • Moore's Lobby · 1h 00m

    Burning Man Maker, Music Mogul, and PCB Revolutionary—Flux CEO Matthias Wagner

    Matthias Wagner, CEO of Flux, discusses his unconventional career path from Siemens to music production and back to hardware, explaining how his experience creating the viral 'Crazy Frog' track and designing modular energy systems for Burning Man informed his vision for Flux. He details the frustration hardware engineers face with legacy, waterfall-style PCB design tools, contrasting them with the fluid, real-time collaboration found in modern software and music production workflows. Wagner explains how Flux aims to democratize hardware design by combining GitHub-style version control with Figma-like usability, allowing users to iterate quickly without managing complex file formats or manual footprints. The episode concludes with a look at Flux's new AI assistant, which helps engineers with component selection, schematic validation, and even generating entire designs from natural language prompts.

  • Embedded.fm · 59m

    464: Please Make This Monster Look Scary

    Elecia White and Christopher White discuss Elecia's upcoming third edition of 'Making Embedded Systems' and Christopher's contract work with Logical Elegance. They debate the design of a potential embedded.fm branded PCB, weighing the utility of a simple radio against the complexity of a full development board, and explore the ethical implications of using Large Language Models for creative and professional tasks. Christopher also provides an update on his progress assembling a QRP Labs QCX+ transceiver for amateur radio communication with his father.

  • Embedded.fm · 31m

    Bonus: From Solid Con

    Elecia White interviews several hardware innovators at O'Reilly's Solid Conference, including Erin Mulcahy of littleBits, who demonstrates magnetic modular electronics and the new Arduino and cloud modules for prototyping. Jack Mudd of Onewheel explains the self-balancing electric skateboard's firmware tuning and successful Kickstarter, while Taylor Stein of AutoDesk highlights the free enthusiast license for Fusion 360 CAD software. The episode also covers Qualcomm's Laurie Yoler on IoT connectivity, Playtabase's Ahmed Daoud on the Remo gesture controller, Panoptes UAV's Terrence McKenna on collision avoidance sensors, and Coversant's Michael Holdmann on XMPP-based IoT service buses.

  • Embedded.fm · 47m

    417: I Don't Know How My Brain Works

    Alexandra Covor, a product manager at Zalmotek and Ada Lovelace Fellow, discusses her journey as a maker who blends electronics with art. She details specific projects like the tiny violin PCB, the PicoLight studio light, and an automated hydroponic strawberry system using Edge Impulse and Waylay. Covor also shares insights on her master's curriculum, the Romanian tech scene, and her philosophy on open-source hardware and PCB design aesthetics.

  • OnTrack · 44m

    Code-Based PCB Design with AI: Inside atopile

    Ioannis Papamanoglou and Narayan Powderly, co-founders of atopile, explain their code-based, AI-driven approach to PCB design, arguing that hardware has lagged behind software in tooling automation. They discuss how capturing engineering intent in code allows for deterministic design and seamless integration with tools like Altium Designer and LTspice, shifting the designer's role from boilerplate tasks to system architecture. The founders highlight the importance of connecting disconnected tools to improve supply chain resilience and firmware integration, while noting that the bottleneck for AI in hardware is orchestration rather than model capability.

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by IoT For All Podcast. The episode itself belongs to its publisher.