Making Mood Changes Visible with AI
At the end of May, a public demonstration took place at the Tampere tram depot, illustrating how passengers’ emotional atmosphere can be detected and visualized using AI and video technology. In the demo, a "disruptive" passenger boarded the tram, quickly altering the mood: expressions became serious, chatter quieted, and the mood indicator color on the display shifted from yellow to red.
This negative change in mood was detected using technology that analyzes facial expressions. The analytics do not identify individuals but rather construct an overall picture of the carriage’s emotional atmosphere, which is relayed to the driver.
“With this information, disruptive behaviour can be detected in advance and the passenger experience improved,” says Ambientia’s Design Director, Tuuli Aalto.
Video Analytics–Based Technology
Currently, passengers’ emotional atmosphere is measured primarily through video cameras and AI. The system aims to recognize basic emotions—such as joy, sadness, frustration, or anger—from facial cues and interpret the emotional atmosphere in real time.
As Teleste’s Product Manager Tiia Järvenpää notes: “For example, distinguishing between a sad and an angry expression is not easy.”
Ambientia’s Role: Turning Data into Usable Insights
Ambientia’s solutions process the emotional atmosphere data into a format useful and understandable to the driver. Ambientia handled the integration—retrieving data from Teleste’s system and converting it into actionable information—and developed a user interface that allows the driver to monitor the mood in real time.
“The goal was to integrate emotion analytics into the driver’s and operator’s situational awareness without overloading the driver with information. The driver’s main task remains to drive the tram safely,” Aalto summarizes.
Human-Centered AI, Ethically Developed
Alongside technological development, the HIPE project has prioritized ethical and regulatory considerations. The system must protect passenger anonymity and privacy under all circumstances.
According to VTT Senior Scientist Sari Järvinen, the collected data could in the future enable predictive capabilities. For example, changes in emotional atmosphere associated with particular times of day, routes, or stops could be detected and addressed proactively.
Tested in Tampere, Potential Across the Industry
The demonstration held in May was conducted in Tampere’s Lyyli test tram, which allows new technology to be trialed in a moving environment. Although Tampere Tram is not officially part of the project, the pilot has generated interest in broader public transport innovation.
The HIPE project continues until November 2025. The next phase will focus on commercialization potential and further development. The project involves several Finnish technology and research organizations, including VTT, Teleste, Ambientia, Framery, Isku, Helvar, and the University of Lapland.
