Part of the 11th ACM/IEEE Symposium on Edge Computing (SEC 2026) Location: Santa Clara University | Date: Oct 13-15, 2026
Overview
The future of the Edge is not just decentralized—it must be sustainable. As AI models grow in complexity, the environmental and energetic costs of deploying them on edge devices have become a critical bottleneck.
The Minimalist Edge AI Challenge invites student teams to tackle the "AI Bloat" head-on. This one-day competition challenges you to take state-of-the-art "heavy" models and transform them into lean, green, high-performance machines capable of running on constrained, eco-friendly hardware. Can you maintain intelligence while slashing the carbon footprint?
The Challenge
Participating teams will be tasked with deploying a complex AI workload (e.g., real-time multi-object tracking or environmental anomaly detection) onto provided low-power hardware (Qualcomm Arduino UNO Q).
Your mission is to optimize for the "Green Trifecta":
-
Model Accuracy: Maintain high predictive performance and reliability while operating under strict resource constraints.
-
Model Compression: Use pruning, quantization, and distillation to shrink the model footprint.
-
Resource Efficiency: Minimize peak RAM and CPU utilization to ensure smooth execution on minimalist hardware.
-
Energy Parsimony: Achieve the lowest possible Joules-per-inference to maximize battery life and sustainability.
Cash Prizes:
- 1st Place: $1,000
- 2nd Place: $750
- 3rd Place: $500
