BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York BEGIN:DAYLIGHT DTSTART:20260308T030000 TZOFFSETFROM:-0500 TZOFFSETTO:-0400 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:EDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20261101T010000 TZOFFSETFROM:-0400 TZOFFSETTO:-0500 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:EST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20260320T143314Z UID:9E64466D-EBD9-45DF-8D25-6A20C59E389C DTSTART;TZID=America/New_York:20260317T190000 DTEND;TZID=America/New_York:20260317T200000 DESCRIPTION:Physical AI at the Edge: Autonomous Drones\, Embedded Vision\, and Hardware-Optimized Learning\n\nEdge AI is redefining how intelligent s ystems operate in real-world\, bandwidth-limited\, and safety-critical env ironments. Rather than relying on centralized cloud infrastructure\, next- generation autonomous platforms demand low-latency\, energy-efficient\, an d hardware-aware\nintelligence deployed directly at the edge. This guest l ecture presents a systems-level perspective on designing edge-native AI ar chitectures spanning three converging domains: autonomous drone platforms\ , embedded computer vision\, and lightweight large language models (LLMs). We will examine how drone autonomy benefits from onboard perception-actio n loops\, real-time multi-modal sensing\, and closed-loop control framewor ks that operate without persistent cloud connectivity.\n\nThe talk further explores edge-optimized vision pipelines for agricultural and environment al monitoring\, including model compression\, quantization\, spectral-RGB fusion\, and real-time deployment on resource-constrained hardware such as Jetson\, FPGA\, and heterogeneous SoCs.\nFinally\, we discuss the emergin g role of LLMs in physical AI systems\, not as chat interfaces\, but as st ructured reasoning engines integrated with robotic sensing and decision pi pelines. Hardware-aware optimization strategies\, including pruning\, mixe d-precision inference\, memory- latency tradeoffs\, and accelerator-centri c design\, will be highlighted as key enablers of scalable deployment. Thi s invited talk concludes by outlining a unifying design framework for buil ding autonomous\, interpretable\, and deployable edge intelligence systems across agriculture\, environmental monitoring\, and cyber-physical domain s.\n\nAll are welcome! You do not need to be a member to attend. If you ar e interested and unable to attend\, please register and a recording will b e sent out after the event.\n\nSpeaker(s): Prabha \, \n\nVirtual: https:// events.vtools.ieee.org/m/531954 LOCATION:Virtual: https://events.vtools.ieee.org/m/531954 ORGANIZER:ieee.lvs.wie@gmail.com SEQUENCE:13 SUMMARY:Women in Sustainability Series 2026 - Physical AI at the Edge: Dr. Prabha Sundaravadivel URL;VALUE=URI:https://events.vtools.ieee.org/m/531954 X-ALT-DESC:Description: <br /><h1><strong>Physical AI at the Edge: Autonomo us Drones\, Embedded Vision\, and Hardware-Optimized Learning</strong></h1 >\n<p>Edge AI is redefining how intelligent systems operate in real-world\ , bandwidth-limited\, and safety-critical environments. Rather than relyin g on centralized cloud infrastructure\, next-generation autonomous platfor ms demand low-latency\, energy-efficient\, and hardware-aware<br>intellige nce deployed directly at the edge. This guest lecture presents a systems-l evel perspective on designing edge-native AI architectures spanning three converging domains: autonomous drone platforms\, embedded computer vision\ , and lightweight large language models (LLMs). We will examine how drone autonomy benefits from onboard perception-action loops\, real-time multi-m odal sensing\, and closed-loop control frameworks that operate without per sistent cloud connectivity.</p>\n<p>The talk further explores edge-optimiz ed vision pipelines for agricultural and environmental monitoring\, includ ing model compression\, quantization\, spectral-RGB fusion\, and real-time deployment on resource-constrained hardware such as Jetson\, FPGA\, and h eterogeneous SoCs.<br>Finally\, we discuss the emerging role of LLMs in ph ysical AI systems\, not as chat interfaces\, but as structured reasoning e ngines integrated with robotic sensing and decision pipelines. Hardware-aw are optimization strategies\, including pruning\, mixed-precision inferenc e\, memory- latency tradeoffs\, and accelerator-centric design\, will be h ighlighted as key enablers of scalable deployment. This invited talk concl udes by outlining a unifying design framework for building&nbsp\;autonomou s\, interpretable\, and deployable edge intelligence systems across agricu lture\,&nbsp\;environmental monitoring\, and cyber-physical domains.</p>\n <p>&nbsp\;</p>\n<p>All are welcome! You do not need to be a member to atte nd. If you are interested and unable to attend\, please register and a rec ording will be sent out after the event.</p> END:VEVENT END:VCALENDAR