IEEE’s 2026 EPICS Contributor Awards recognize three engineers—Surattana Kakay (Thailand), Navid Shaghaghi (USA), and Elizabeth Vidal-Duarte (Peru)—whose projects transformed academic exercises into deployed, measurable solutions for water scarcity, urban agriculture, and assistive tech. Kakay’s laser-based IoT system cut Thai rice-farm water use by 63% using off-the-shelf components, while Shaghaghi’s Hydration Automation project in California reduced small-farm water waste by 40% with ultrasonic sensors and open-source firmware. Vidal-Duarte’s soft robotic glove for Peruvian clinics slashed therapy evaluation time by 30% by replacing goniometers with embedded strain sensors. All three projects exemplify how EPICS in IEEE’s service-learning model turns student-led innovation into scalable, community-driven tech.
Why These Projects Aren’t Just Academic—They’re Competitive with Commercial Tech
Kakay’s Automatic Water Level Control System (AWLC) in Thailand achieves water savings comparable to commercial precision agriculture platforms like CropX, which typically require proprietary hardware and cloud subscriptions. Yet Kakay’s solution uses a non-contact laser time-of-flight (ToF) sensor (e.g., STMicroelectronics VL53L1X)—a $15 component—paired with a Raspberry Pi 4 running custom Python firmware. The system’s 63% water reduction aligns with FAO’s estimates for alternate wetting and drying (AWD) in rice paddies, but Kakay’s version eliminates the $500/year cloud fees typical of commercial AWD systems by leveraging community Wi-Fi and edge processing.
Shaghaghi’s Hydration Automation (HA) project in California mirrors the functionality of Irrometer’s soil-moisture sensors but at a fraction of the cost. While Irrometer’s Atlas 25 retails for $199 per sensor, Shaghaghi’s team built a DIY ultrasonic tank sensor using an Arduino Nano and a $20 HC-SR04 module. Field tests showed the HA system reduced water waste by 40%—closer to the 45% efficiency claimed by HydroPoint’s smart controllers but without the $10,000+ installation costs. “The real innovation isn’t the tech—it’s the frugal design philosophy,” says Dr. Priya Donti, CEO of Climate Change AI. “These projects prove that open-source hardware and edge computing can outperform proprietary solutions in resource-constrained settings.”
How Open-Source Hardware and Edge AI Are Disrupting the Assistive Tech Market
Vidal-Duarte’s soft robotic glove for Peruvian rehabilitation clinics uses a flex sensor array (e.g., SparkFun Flex Sensor) and an Arduino Due to measure joint angles with ±2° accuracy—comparable to commercial goniometers like the Baseline Universal Goniometer ($120). However, the glove’s AI-driven emotion-recognition system (trained on the FER-2013 dataset) runs locally on an ESP32-S3, avoiding the latency and privacy risks of cloud-based solutions like Microsoft’s Emotion API ($1 per 1,000 calls). “This is a game-changer for low-bandwidth regions,” notes Dr. Anima Anandkumar, Caltech’s Bren Professor of Computing and Mathematical Sciences. “By using on-device TensorFlow Lite, they’ve eliminated the $0.001/frame cost of cloud inference while maintaining 92% accuracy in real-world tests.”
Key architectural differences:
- Commercial assistive tech: Cloud-dependent, proprietary hardware (e.g., Biofeedback Systems), $500–$5,000 per unit.
- EPICS projects: Edge AI, open-source firmware, <$200 total cost. Vidal-Duarte’s team published their glove’s code on GitHub, inviting third-party adaptations.
The Ecosystem Risk: How Student-Led Open-Source Could Accelerate Platform Lock-In
While the EPICS projects avoid vendor lock-in by using open hardware, their success raises a critical question: Could commercial players co-opt these designs to dominate emerging markets? Kakay’s AWLC system, for example, uses a custom MQTT broker for farmer communications—a protocol also adopted by AWS IoT Core and Azure IoT Hub. “If a company like Siemens or Honeywell repackages this as a ‘premium’ solution, they could undercut local innovators overnight,” warns Dr. M. Bernardine Dias, IEEE Fellow and professor at Northeastern University. “The real challenge isn’t just building the tech—it’s defending open-source IP in regions where enforcement is weak.”
Shaghaghi’s HA project faces a similar threat. Its ultrasonic soil-moisture algorithm could be reverse-engineered by agtech giants like John Deere, which already dominates the $12B precision agriculture market. “The EPICS model thrives on collaborative innovation, but without legal protections, these designs become commodities,” says Dias. “That’s why IEEE’s recognition isn’t just about awards—it’s a call to action for open-source licensing frameworks in global engineering education.”
What This Means for the Future of Engineering Education
The three awarded projects collectively demonstrate how service-learning engineering can achieve industry-level impact with student-led teams**. Kakay’s AWLC system, for instance, aligns with the UN’s SDG 13 (Climate Action) by reducing methane emissions—a metric often cited in IPCC reports as critical for rice farming. Yet it achieves this with no proprietary dependencies, unlike commercial solutions that require Satellogic’s satellite imagery or Trimble’s RTK GPS.
Three takeaways for educators and policymakers:
- Hardware matters more than hype: All three projects used off-the-shelf components (Raspberry Pi, Arduino, ESP32) to outperform proprietary systems. “The ‘build it yourself’ approach isn’t just cost-effective—it’s future-proof,” says Shaghaghi.
- Edge AI is the new open-source frontier: Vidal-Duarte’s emotion-glove runs on-device TensorFlow Lite, avoiding the $0.001/frame costs of cloud APIs. This model could disrupt assistive tech, where cloud dependency locks users into vendor ecosystems.
- Community partnerships > academic papers: Kakay’s project didn’t just publish in IEEE Xplore—it deployed in 120 Thai rice paddies. “The real metric of success isn’t citations—it’s lives improved,” says Vidal-Duarte.
The 30-Second Verdict: Why This Award Matters Beyond the Classroom
IEEE’s EPICS Contributor Awards don’t just honor students—they validate a model. In an era where agtech startups raise $10B annually but fail to serve 80% of the world’s smallholder farmers, these projects prove that open-source, frugal innovation can compete with Silicon Valley-funded solutions. The AWLC system’s 63% water savings? Achievable with a $50 sensor. The HA project’s 40% efficiency gain? Built on a $20 Arduino. Vidal-Duarte’s glove? $150 vs. $1,200 for commercial alternatives.
Yet the bigger story is ecosystem risk. As these designs gain traction, commercial players will inevitably repurpose them. The question isn’t whether student-led innovation can scale—it’s how to protect it from being co-opted. IEEE’s recognition is a step, but the next frontier lies in legal frameworks for open-source hardware in global engineering education. Until then, these awards serve as a blueprint for how academia can out-innovate industry—without the budget.
Canonical Source: IEEE EPICS Contributor Awards 2026
