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:20260717T011121Z UID:620D8F68-0030-404E-8711-CF647A0B5191 DTSTART;TZID=America/New_York:20260721T183000 DTEND;TZID=America/New_York:20260721T210000 DESCRIPTION:Miniature robots are useful during disaster response and access ing remote or unsafe areas. They need to navigate uneven terrains without supervision and under severe resource constraints such as limited compute\ , storage and power budget. Biological locomotion relies on a hierarchical structure where Central Pattern Generators (CPGs) provide rhythmic primit ives that are modulated by descending signals to produce complex motor pat terns. Event-based sensorimotor control in edge robotics has potential to enable fully autonomous and adaptive robot navigation systems capable of r esponding to environmental fluctuations by learning new types of motion an d real-time decision making to avoid obstacles.\n\nIn this talk\, I will p resent my team’s work on a novel bio-inspired framework with a hierarchi cal control system\, utilizing a tunable multi-layer neural network with a hardware-friendly CPG as the core coordinator to govern the precise timin g of periodic motion. Autonomous operation is managed by a Dynamic State M achine (DSM) at the top of the hierarchy\, providing the necessary adaptab ility to handle environmental challenges such as obstacles or uneven terra in. The multi-layer neural network uses a nonlinear neuron model which emp loys mixed feedback at multiple timescales to produce rhythmic patterns of bursting events to control the motors. The system was verified virtually in Gazebo simulation and validated in the real world via deployment on a R aspberry Pi platform. My team will demonstrate the proposed framework on t he Petoi robot\, which can autonomously learn walk and crawl gaits using s upervised Spike-Time Dependent Plasticity (STDP) and surrogate gradient le arning algorithms\, transition between the learned gaits stored as new sta tes\, through the DSM for real-time gesture controlled navigation and obst acle avoidance. I will conclude with experimental results that demonstrate real-time neuromorphic control in low-power robotic systems.\n\nSpeaker(s ): \, Rajkumar Kubendran\n\nAgenda: \n6:30 PM Gather and meal\n\n7:00 PM Presentation\n\n8:30 Adjourn\n\nRoom: 102\, Bldg: Benedum Hall\, 3700 O'Ha ra Street\, Pittsburgh\, Pennsylvania\, United States\, 15213\, Virtual: h ttps://events.vtools.ieee.org/m/560566 LOCATION:Room: 102\, Bldg: Benedum Hall\, 3700 O'Hara Street\, Pittsburgh\, Pennsylvania\, United States\, 15213\, Virtual: https://events.vtools.iee e.org/m/560566 ORGANIZER:rsprang@ieee.org SEQUENCE:27 SUMMARY:End-to-end Neuromorphic Sensorimotor Control with Supervised Gait L earning for Autonomous Robot Navigation URL;VALUE=URI:https://events.vtools.ieee.org/m/560566 X-ALT-DESC:Description: <br /><p style="line-height: 116%\; margin-bottom: 0.11in\;" align="justify">Miniature robots are useful during disaster resp onse and accessing remote or unsafe areas. They need to navigate uneven te rrains without supervision and under severe resource constraints such as l imited compute\, storage and power budget. Biological locomotion relies on a hierarchical structure where Central Pattern Generators (CPGs) provide rhythmic primitives that are modulated by descending signals to produce co mplex motor patterns. Event-based sensorimotor control in edge robotics ha s potential to enable fully autonomous and adaptive robot navigation syste ms capable of responding to environmental fluctuations by learning new typ es of motion and real-time decision making to avoid obstacles.</p>\n<p sty le="line-height: 116%\; margin-bottom: 0.11in\;" align="justify">In this t alk\, I will present my team&rsquo\;s work on a novel bio-inspired framewo rk with a hierarchical control system\, utilizing a tunable multi-layer ne ural network with a hardware-friendly CPG as the core coordinator to gover n the precise timing of periodic motion. Autonomous operation is managed b y a Dynamic State Machine (DSM) at the top of the hierarchy\, providing th e necessary adaptability to handle environmental challenges such as obstac les or uneven terrain. The multi-layer neural network uses a nonlinear neu ron model which employs mixed feedback at multiple timescales to produce r hythmic patterns of bursting events to control the motors. The system was verified virtually in Gazebo simulation and validated in the real world vi a deployment on a Raspberry Pi platform. My team will demonstrate the prop osed framework on the Petoi robot\, which can autonomously learn walk and crawl gaits using supervised Spike-Time Dependent Plasticity (STDP) and su rrogate gradient learning algorithms\, transition between the learned gait s stored as new states\, through the DSM for real-time gesture controlled navigation and obstacle avoidance. I will conclude with experimental resul ts that demonstrate real-time neuromorphic control in low-power robotic sy stems.<span lang="en-IN">&nbsp\;</span></p><br /><br />Agenda: <br /><p>6: 30 PM Gather and meal</p>\n<p>7:00 PM Presentation</p>\n<p>8:30 Adjourn</p > END:VEVENT END:VCALENDAR