BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//TECoSA - ECPv6.16.3//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:TECoSA
X-ORIGINAL-URL:https://www.tecosa.center.kth.se
X-WR-CALDESC:Events for TECoSA
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:UTC
BEGIN:STANDARD
TZOFFSETFROM:+0000
TZOFFSETTO:+0000
TZNAME:UTC
DTSTART:20220101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=UTC:20231005T150000
DTEND;TZID=UTC:20231005T160000
DTSTAMP:20260613T021113
CREATED:20230817T102203Z
LAST-MODIFIED:20230817T102238Z
UID:6775-1696518000-1696521600@www.tecosa.center.kth.se
SUMMARY:TECoSA Seminar - Configuration of Dependable Edge Computing Platforms for Virtualized Critical Control Applications
DESCRIPTION:We aim to bring you a TECoSA Seminar on the first Thursday of each month during term-time.  All are welcome to attend and we look forward to some lively discussions. Members can accept the Outlook invitations\, non-members can email tecosa-admin@kth.se to register.\nOur October seminar is with Prof Paul Pop\, Head of the Embedded Systems Engineering section at DTU Compute\, DTU Technical University of Denmark.  The session will be given via Zoom (https://kth-se.zoom.us/j/66857695267). \nABSTRACT: Edge Computing offers a unified platform where applications of varying criticality coexist. Critical control applications\, characterized by their periodic hard real-time tasks and stringent timing and safety requisites\, share the same platform with Edge applications which are aperiodic and non-critical. As we advance towards Industry 4.0 and embrace Time-Sensitive Networking (TSN)\, the integration of traditional systems like Programmable Logic Controllers (PLCs) with cyber-physical systems via virtualization is paramount. The talk presents the problem of configuring of dependable Edge Computing Platforms (ECPs) for virtualized critical control applications. The configuration poses an optimization challenge aiming to harmonize critical control and Edge applications. We discuss solutions to the placement of virtual PLCs in ECPs and the routing of TSN traffic in industry-based networks\, as well as solutions to facilitate an ECP’s extensibility for adding future control applications without exhaustive re-certification\, also supporting the hosting of multiple dynamic Edge applications. \nBIO:  Paul Pop is a Professor of Cyber-Physical Systems at DTU Compute\, Technical University of Denmark (DTU). He has received his Ph.D. degree in computer systems from Linköping University in 2003. His research is focused on developing methods and tools for the analysis and optimization of networked dependable cyber-physical systems. In this area\, he has published over 150 peer-reviewed papers\, three books\, and seven book chapters. He has served as a technical program committee member on several conferences\, such as DATE and ESWEEK. He has received the Best Paper Award at DATE 2005\, RTIS 2007\, CASES 2009\, MECO 2013\, DSD 2016\, ETFA 2020\, and an outstanding paper award at RTNS 2022. He is the coordinator of the Nordic University Hub on Industrial IoT has coordinated the European Training Network on Fog Computing for Robotics and Industrial Automation.
URL:https://www.tecosa.center.kth.se/event/tecosa-seminar-connfiguration-of-dependable-edge-computing-platforms-for-virtualized-critical-control-applications/
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=UTC:20231027T120000
DTEND;TZID=UTC:20231027T130000
DTSTAMP:20260613T021113
CREATED:20231017T070710Z
LAST-MODIFIED:20231109T131300Z
UID:6892-1698408000-1698411600@www.tecosa.center.kth.se
SUMMARY:TECoSA Research Seminar: Getting the Best Out of Both Worlds: Algorithms for Hierarchical Inference at the Edge
DESCRIPTION:Speaker: Vishnu Narayanan Moothedath\, TECoSA PhD student\nVenue\, Zoom link and sign-up link circulated to members\nPlease email vickid@kth.se if you have any questions. \nABSTRACT: We consider a resource-constrained Edge Device (ED)\, such as an IoT sensor or a microcontroller unit\, embedded with a small-size ML model (S-ML) for a generic classification application\, and an Edge Server (ES) that hosts a large-size ML model (L-ML). Since the inference accuracy of S-ML is lower than that of the L-ML\, offloading all the data samples to the ES results in high inference accuracy\, but it defeats the purpose of embedding S-ML on the ED and deprives the benefits of reduced latency\, bandwidth savings\, and energy efficiency of doing local inference. In order to get the best out of both worlds\, i.e.\, the benefits of doing inference on the ED and the benefits of doing inference on ES\, we explore the idea of Hierarchical Inference (HI)\, wherein S-ML inference is only accepted when it is correct\, otherwise the data sample is offloaded for L-ML inference. However\, the ideal implementation of HI is infeasible as the correctness of the S-ML inference is not known to the ED. We thus propose an online meta-learning framework that the ED can use to predict the correctness of the S-ML inference. In particular\, we propose to use the probability corresponding to the maximum probability class output by S-ML for a data sample and decide whether to offload it or not. The resulting online learning problem turns out to be a Prediction with Expert Advice (PEA) problem with continuous expert space. We consider two scenarios\, a full feedback scenario\, where the ED receives feedback on the correctness of the S-ML once it accepts the inference\, and a no-local feedback scenario. We propose the HIL-F and HIL-N algorithms and prove that both of them has sublinear regret bounds without any assumption on the smoothness of the loss function. We evaluate and benchmark the performance of the proposed algorithms for image classification application using different datasets.
URL:https://www.tecosa.center.kth.se/event/tecosa-research-seminar-getting-the-best-out-of-both-worlds-algorithms-for-hierarchical-inference-at-the-edge/
CATEGORIES:Seminar
END:VEVENT
END:VCALENDAR