inMotion
Integrated AI Platform for Urban Mobility and Transportation
An R&D project led by Wavecom in collaboration with Instituto de Telecomunicações (IT Aveiro). Funded by CENTRO 2030 (CENTRO2030-FEDER-02225900).
What we do
Public transport operators lack detailed data on how passengers actually move through the network. Ticketing systems tell you that someone bought a pass, not where they got off. Infrared counters and cameras are expensive, inaccurate in real conditions, and raise privacy concerns.
inMotion develops non-intrusive sensing using Wi-Fi and Bluetooth signals already present on modern buses. By analysing signal strength patterns over time, we infer boarding, alighting, and movement — without cameras, without identifying individuals, and without requiring passengers to install anything.
Two core products
PPS1 — Passenger Detection & Tracking
Anonymous data collection through Wi-Fi and Bluetooth sensing, combined with sensor fusion. Detects how many people board and alight at each stop, feeding origin-destination models.
PPS2 — AI Platform for Urban Mobility
Machine learning and NLP layer for route optimisation, demand prediction, and operator-facing analytics. Designed to let transit managers query data in natural language through RAG (Retrieval-Augmented Generation).
Latest
Our work on RSSI-based passenger movement classification was accepted at EuCNC & 6G Summit 2026. We evaluated 38 classifiers with Bayesian hyperparameter optimisation using Optuna, achieving MCC 0.907 in clean conditions and 0.770 in realistic noisy scenarios. See publications →