BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260730T152640Z
LOCATION:Exhibit Hall
DTSTART;TZID=America/Los_Angeles:20260728T174500
DTEND;TZID=America/Los_Angeles:20260728T174600
UID:dac_DAC 2026_sess306_LBR097@linklings.com
SUMMARY:Late-Breaking Results: Ultra Energy Efficient Personalized Glucose
  Prediction with RLS Implementation on a Tiny FPGA
DESCRIPTION:Georgios Mentzos (Karlsruhe Institute of Technology), Theodora
  Podimata (University Of Patras), Georgios Zervakis (National Technical Un
 iversity of Athens), Emmanouil Psarakis (University of Patras), and Joerg 
 Henkel (KIT)\n\nAccurate prediction of future blood glucose (BG) levels is
  critical for\nthe effective management of type 1 diabetes. Existing appro
 aches\nfor glucose prediction often rely on deep learning models that are\
 ncomputationally intensive. In this work, we propose an adaptive and\nligh
 tweight recursive least squares (RLS)-based BG predictor that\nenables onl
 ine model updates while maintaining low computational\ncomplexity. Our RLS
  implementation on a tiny FPGA delivers 1.55\n𝜇s latency and 51.3 𝜇J energ
 y consumption per prediction. Evalua-\ntion on the OhioT1DM dataset demons
 trates an RMSE of 18.83 and\n32.12 for 30 and 60 minute prediction horizon
 s, respectively. Our\nFPGA BG predictor outperforms the state-of-the-art (
 SoA) far-edge\nimplementations across all metrics: latency, energy, and er
 ror.\n\nTrack: Student\n\n
END:VEVENT
END:VCALENDAR