Intermittent computing, embedded security, and cyber-physical systems, building batteryless, secure and intelligent systems for the Global South.
Systems that survive power loss mid-execution on batteryless hardware. Our CheckMate work (ACM SenSys 2025) uses LLMs to automate energy-accuracy trade-offs on energy-harvesting devices.
Side-channel attacks and defences on constrained IoT hardware. Our AsiaCCS 2025 paper demonstrates eavesdropping via temporal IMU misalignment on commodity smartphones.
Deployable, zero-maintenance sensing for real infrastructure. Cleanify (NRPU 2024–26) achieves 99.3% LoRaWAN reliability across 5 LUMS campus sites with no batteries.
Consent-driven privacy for wearables and smart glasses. Two papers accepted at IEEE PerCom 2026 and ACM CHI 2026 on bystander privacy with consent-based restoration.
LLM-powered approximate intermittent computing, enabling intelligent energy-accuracy trade-offs on batteryless devices with no hardware overhead.
City-scale batteryless smart waste management. 99.3% reliability across 5,945 real-world lid actuations across 5 LUMS campus deployment sites.
Two 2026 papers extend our bystander-privacy work for smart glasses: PerCom introduces consent-driven privacy by default, and CHI adds consent-based restoration that preserves social context.
View Story →