Democratizing Physical AI with Arduino

For the past two decades, if you wanted to bridge the gap between software and the physical world, you had Arduino. Arduino democratized electronics. But as the industry aggressively pivots towards AI, the barrier to entry has skyrocketed again. Suddenly, building a device capable of running large AI workloads seems to require a PhD in computer science, a team of data scientists, and a massive budget.

During this keynote at the IoT Stars Nuremberg event, Marcello Majonchi from Arduino took the stage to announce that the company is officially tearing that barrier back down.

Fresh off joining the Qualcomm family and shipping a staggering 250,000 units of the dual-brain Arduino UNO Q in just three months, Arduino is making its biggest play yet in the AI space. Here is how they plan to make "Physical AI" accessible to everyone.

Software First: The "Lego Bricks" of AI

It might seem strange for a company famous for its iconic blue PCBs, but Marcello made it clear that Arduino’s primary focus right now is software.

To solve the complexity of Edge AI, Arduino introduced the App Lab, a software development platform designed to abstract away the learning curves of machine learning and embedded infrastructure. The core concept? "Bricks."

"The building blocks of our App Lab are called 'bricks.' You need a database? You have a brick for that. You need an AI model from Edge Impulse? You have a brick for that. You build the brick so you can reuse it, and if you share it, the community will be able to adopt it."
— Marcello Majonchi, Arduino

Instead of writing thousands of lines of bespoke code just to get a sensor talking to an inference engine, developers can snap together these software components like Legos. It’s an iteration of the old Arduino Uno libraries, a completely new ecosystem designed to reduce the friction of the developers creativity in the AI era.

Meet the Ventuno Q: 40 TOPS of Physical AI

To run this new software paradigm, you need serious silicon. Enter the Arduino VENTUNO Q. Named "Ventuno" (Italian for 21) to celebrate the 21st anniversary of the Arduino foundation, this new board brings cutting-edge AI capabilities to the embedded developer's workbench. Powered by Qualcomm silicon iQ8, the VENTUNO Q brings 40 TOPS (Trillions of Operations Per Second) of computational power.

But Marcello was quick to distinguish what Arduino is doing from standard edge computing. They aren’t just building an AI processor; they are building a platform for Physical AI.

"With this VENTUNO Q, we are really catering to developers that want to create physical AI applications, applications that bring together AI and interaction with the physical world, whether it is collecting data from a sensor or affecting change by controlling a motor or a robot."

The Real-World Difference

During the Q&A, IoT Stars audience challenged Marcello with the question every maker was thinking:

“I have a Raspberry Pi, a Jetson Nano, and now a VENTUNO Q. When do I use what?”

Marcello’s answer perfectly encapsulated Arduino's superpower. While platforms like the Jetson are great for pure computational heavy lifting, they struggle when it comes to the reality of hardware actuation.

"Try to connect a motor to a Raspberry Pi or a Jetson Nano. Then try to connect it with a VENTUNO Q and see what your AI can actually affect in the real world. You'll see the difference between the two."

Open Source by Design

Despite the massive leap in processing power and the partnership with Qualcomm, Arduino remains fiercely committed to its roots. Marcello closed the keynote with a promise to the developer community: everything they are building for the Edge AI era remains fully open source.

Developers can buy the VENTUNO Q, prototype a physical AI system, and the moment they are ready to scale to production, they can take the open-source hardware design and software stack, modify it, and ship it with no strings attached.

For 21 years, Arduino made sure anyone could blink an LED. Now, they are making sure anyone can build an autonomous AI-driven machine.