DigiKey Demos Prototyping at embedded world

Rapid prototyping was the main thread through Elektor’s visit to DigiKey at embedded world 2026, where I David Sandys, Senior Director, Technical Marketing, about a booth full of practical engineering demos. As part of Elektor’s wider embedded world 2026 coverage, the conversation showed how development boards, sensors, robotics, measurement tools, AI hardware, and connector ecosystems are changing the way engineers get from idea to working prototype.

Rapid Prototyping on the DigiKey Booth

The first demo centered on the Experiential Robotics Platform, or XRP, a small robotics platform developed with partners including Raspberry Pi, SparkFun, FIRST Robotics, and WPI. Sandys used a Red Pitaya as a digital oscilloscope to show what happens electrically when the XRP’s sensor is covered. Instead of only seeing a robot respond, students can see the voltage levels and square waveforms behind that response.
 

DigiKey’s own applications engineering team built the board used in the demo, which fits the wider point Sandys made about the XRP. Low-end robotics kits are often useful, but they can stop at assembly and basic coding. XRP was designed to be expanded, which gives educators and makers a better route from classroom robotics into real hardware investigation. Pairing it with Red Pitaya also connects robotics to test and measurement rather than treating those as separate worlds.

Rapid Prototyping with AI, FPGA, and Connector Ecosystems

The walk-through then moved to an Altera DE25 board from Terasic, where video imaging was handled in the FPGA in real time — no Linux, no boot time. He then showed me a demo using the STM32N6 for hand-gesture recognition. In that setup, gestures such as a thumbs-up, number signs, and a pinch gesture were used to select and control simple games. It was a compact example of edge AI as a real interface technique.

Arduino also featured in the conversation. Sandys showed the UNO Q and the newer VENTUNO Q, pointing out features such as an M.2 connector, Raspberry Pi interface, three camera-module connections, and a Qualcomm SOM boasting 40 TOPS. One demo used the board for real-time person recognition, while another combined UNO Q with Machinechat for local data aggregation and visualization.

It’s useful that the data is displayed, but the fact that it remains on-premises is significant tor users who do not want every temperature value or sensor stream sent to AWS or another cloud platform and require an always-on internet connection.

Sandys also described the shift from individual components to complete ecosystems. Components remain DigiKey’s core business, but engineers are increasingly building rapid prototypes around Qwiic, STEMMA, Grove, and MikroElektronika Click connectors. Sandys made the point that what used to be measured in days can now often be measured in hours. A student, maker, or engineer can order boards and modules, connect them without designing every interface from scratch, and get to a working proof of concept much faster.

From Beagle Zepto to a Thread Network

The Beagle ecosystem came up several times. Sandys showed the BeagleBadge and highlighted the Beagle Zepto, a roughly $1 TI MSPM0-based board with an RGB LED and MikroElektronika Click connector support. He also pointed out that the Zepto could be powered through a Qwiic connector by another Beagle platform, which is exactly the sort of small hardware detail that matters when people are trying to build quickly.

The most playful demo used XRP as the basis for AI playing AI at chess. Two SparkFun motor drivers and two stepper motors moved the pieces, while the pieces themselves were built from electronic components: potentiometers as pawns, fuses and capacitors elsewhere, MOSFETs and transistors as queens, and large capacitors with a 555 timer as the king’s crown. It is a booth demo, yes, but it is also a good explanation of modular design: mechanics, control, software, motors, and electronics all have to cooperate before the trick works.

The final demo pulled several vendors into one system. An NXP Freedom platform acted as the Thread router, a Microchip Curiosity board controlled a motor and fan, another Freedom platform acted as a Thread device, and an STM32 board joined the same network, with Home Assistant coordinating the system.

Customers rarely build everything around one supplier’s silicon. They choose the right technology for each part of the application, and the hard part is often making those parts work together. For modern rapid prototyping, that integration layer is where much of the real engineering now happens.