Last month, the first edition of the European School of Information Theory Data-to-Information (D2I) Competition took place in conjunction with ESIT 2026 in Nordfjordeid, Norway. This student competition, organised by the IEEE Information Theory Society Student and Outreach Committee, brought together data science, machine learning and information theory around a real-world challenge: working with Channel State Information (CSI) measurements to improve wireless localisation under noisy and imperfect conditions.
This year’s challenge was inspired by the Viking legend of Bifröst, the shimmering bridge connecting Midgard to Asgard. While reaching Asgard once required divine intervention, a winged horse and, ideally, a heroic death in battle, today’s Nordic explorers benefit from 5G coverage reaching roughly 99.7% of households in Norway. The question posed to participants was therefore simple, but far from easy: given noisy, partial and imperfect measurements of the wireless channel, could they chart their way to Bifröst?
Next week, the selected finalist teams will present their approaches and results in an online session supported by ELLIS Unit Madrid. Their presentations will focus on channel charting, robust learning and unlearning noisy data contributions, with the final goal of improving localisation quality from real-world CSI measurements.
All interested participants are warmly invited to attend the finalist presentation session, sponsored by ELLIS Unit Madrid. The session will take place online on 7 July, from 16:00 to 17:00 CEST (GMT+2), via Google Meet: https://meet.google.com/vpv-bsye-wyv