BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//denog15//speaker//9KC9DC
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-denog15-8AN7PV@pretalx.com
DTSTART;TZID=CET:20231120T143000
DTEND;TZID=CET:20231120T150000
DESCRIPTION:Communication networks are vital for society and network availa
 bility is therefore crucial. There is a huge potential in using network te
 lemetry data and machine learning algorithms to proactively detect anomali
 es and remedy network problems before they affect the customers. In practi
 ce\, however\, there are many steps on the way to get there. In this prese
 ntation we would like to share the status of an ongoing research collabora
 tion with the purpose of simplifying the operation and increase the availa
 bility of fiber access networks.\n\nThe research project is part of a CELT
 IC_NEXT flagship research program (AI-NET) that has the overall target of 
 accelerating the digital transformation in Europe by intelligent network a
 utomation.\n\nThe project (“Palantir”) consists of two parts:\n\nA fie
 ld trial in a Swedish municipality network where telemetry data from more 
 than 500 access switches\, connecting more than 12000 households\, are col
 lected and analyzed for anomalies using machine learning. (The field trial
  is planned to be expanded to cover 1000 access switches during the projec
 t and synthetic errors are planned to be injected to emulate error situati
 ons.)\n\nThe second part is a demonstrator to be set up at Fraunhoffer HHI
  in Berlin (starting in October 2023). The demonstrator covers an end to e
 nd network and aims to demonstrate all aspects of the research program.\n\
 nIn the presentation we would like to describe our field trial and demonst
 rator\, share our experiences in collecting and analyzing telemetry data i
 n the field and describe our conclusions so far. \nWe would also like to o
 pen up for a discussion with the DENOG15 participants on what real life er
 ror situations that causes the most problems in fiber access networks and 
 whether they would be suitable for AI detection (and potentially also for 
 error emulation in our field trial).\n\nProject Palantir is a collaboratio
 n between:\n•	Fraunhofer HHI	 (German Research Institute): Demonstrator\
 n•	Lunet (Swedish operator - open access municipality network): Field tr
 ial\n•	RISE (Independent Swedish research institute): Field trial\n•	S
 avantic (Swedish AI specialists): Field trial\n•	Waystream (FTTX vendor)
 : Field trial and demonstrator
DTSTAMP:20260413T210102Z
LOCATION:Auditorium
SUMMARY:Towards automated and proactive anomaly detection in fiber access n
 etworks - Johan Sandell\, Behnam Shariati\, Christian Burk\, Dr. David Hoc
 k
URL:https://pretalx.com/denog15/talk/8AN7PV/
END:VEVENT
END:VCALENDAR