At India AI Impact Summit, “AI for All” and student innovation take center stage 

On February 18th, Arduino joined the India AI Impact Summit 2026 at Sushma Swaraj Bhawan with a clear message: artificial intelligence isn’t just for well-funded research labs and big tech companies. When you put powerful, accessible tools in the hands of students and educators, remarkable things happen!

To prove it, Arduino and Qualcomm launched an AI Innovation Challenge ahead of the event – and received over 40 projects submitted by students from 20 top universities across India. Each one addressed real-world problems in healthcare, agriculture, education, energy, public services, and more. These weren’t hypothetical exercises or classroom demos, but working prototypes built on the Arduino® UNO™ Q board, tackling challenges that matter to real people in the real world.

Making AI accessible

Our keynote session, “AI for All: From Human Potential to Global Impact,” focused on what happens when you democratize access to edge AI technology. Fabio Violante, VP & GM of Arduino at Qualcomm Europe, spoke about how open ecosystems enable innovation at scale, while Guneet Bedi, Senior Director of Sales at Qualcomm Technologies, outlined our ambitious impact targets across K-12 education, higher education, and industry in India.

The session also included announcements of major AI initiatives and the upcoming launch of our first two “Edge AI on UNO Q” books (arriving May-June 2026). But the real stars of the event were the student teams who took part in the AI Innovation Challenge with projects that demonstrated exactly what “AI for all” looks like in practice. Let’s find out more about three projects that stood out for their technical execution, tangible impact, and innovative use of edge AI capabilities on UNO Q!

1st place: SafeGuard AI – Intelligent fall detection for elderly safety

Team: Ashish Srivastava, Priyanshi Naghera, Dharm Saliya, and Martin Shah (Adani University)

Instead of relying on simple threshold sensors that trigger false alarms, the SafeGuard AI team built a tinyML neural network that runs locally on UNO Q to distinguish between normal daily activities and actual falls. The system analyzes motion patterns in real-time using data from an Arduino® Modulino™ Movement sensor.

When a fall is detected, the device locks into emergency mode, sounds an alarm, and – if not cancelled within 10 seconds – uses the Twilio API to make an automated voice call to a caregiver, relaying vital information like impact force and skin temperature.

The architecture takes full advantage of the dual-core design of UNO Q: the MCU handles strict real-time sensor polling at 100 Hz while the processor core runs TensorFlow Lite inference, hosts a web dashboard, and manages cloud API integration. The system even includes a smart calibration mode that learns the user’s walking style to prevent false alarms for active users. Everything runs locally on the device, ensuring user privacy and zero latency. Check out the details on Arduino Project Hub.

2nd place: Stick.AI – Smart walking stick

Team: Aman Gupta and Madhav Gupta (IIT Delhi)

The Stick.AI team reimagined a traditional walking aid as an intelligent assistive system that monitors both safety and health without adding burden to the user. Two sensors are embedded in a custom-designed ergonomic handle: an IMU for motion analysis and a MAX30102 PPG Heart Rate sensor for cardiovascular monitoring.

Rather than threshold-based fall detection, the system uses AI-based motion classification to analyze movement patterns and identify abnormal behavior or potential fall scenarios. The PPG sensor passively collects heart rate and heart rate variability data while the user naturally grips the stick – no wearable straps or chest sensors required.

The software architecture splits responsibilities cleanly: low-level firmware on UNO Q handles time-critical sensor acquisition, a Python gateway layer performs AI inference and signal processing, and a Flask server provides web visualization and long-term data storage. The result is predictive safety and invisible health monitoring built into something people already use every day. Check out the details on Arduino Project Hub.

3rd place: AI-enabled helmet detection safety interlock system

Team: G Shabarish Setty, Sri Ram Gupta, and Sri Krishna R (NIT Calicut)

Road accidents involving two-wheelers are a major cause of fatalities in developing countries, with helmet non-compliance being a primary factor in severe injuries. This team built a safety interlock system that prevents a motorcycle from starting unless the rider is wearing a helmet.

A camera mounted facing the rider captures real-time images when the ignition key is inserted. A computer vision model running on UNO Q detects whether a helmet is present. If yes, a relay closes and allows ignition. If not, a buzzer activates and the relay stays open, preventing the bike from starting.

The system is low-cost, real-time, and doesn’t rely on enforcement or human intervention. It’s a practical solution designed for mass deployment that directly supports SDG 3: Good Health and Well-Being. Check out the details on Arduino Project Hub.

Focus on solving problems, not fighting toolchains

These three projects – and the dozens of others submitted – demonstrate something important: when students have access to tools that combine real-time embedded control with edge AI capabilities, they don’t just learn theory. They build solutions to problems they see in their own communities. The dual-brain architecture in UNO Q combined with Arduino® App Lab makes it possible to prototype sophisticated AI systems without needing expensive infrastructure or specialized expertise. 

We are grateful to all the students who participated in the AI Innovation Challenge, because they show us what’s possible. For more inspiration and innovation, check out Arduino Project Hub: all submitted projects will be published soon, creating a resource for students, educators, and makers across India and beyond. 

Arduino, UNO, and Modulino are trademarks or registered trademarks of Arduino S.r.l.