AI camera system helped keep record Oktoberfest orderly
Tuesday 6th October 2026 on 11:15 in
Bavaria
The 2026 Oktoberfest drew a record 7.4 million visitors without the dangerous overcrowding seen the previous year, BR reported. The city says its new crowd-monitoring system helped manage visitor flows, although main entrances were temporarily closed at times on the final two Saturdays.
The festival attracted 900,000 more people than the year before. According to the festival management, the grounds were never overcrowded. Munich’s head of labor and economic affairs, Christian Scharpf, said the revised safety plan had worked well. A detailed evaluation is due in November.
The plan combined crowd spotters on the grounds, a new security centre in the Theresienwiese service centre and AI-assisted video monitoring. More than 50 cameras at 35 locations tracked crowd density and movement in real time. The information was displayed across about 22 square metres of screens at the security centre.
Crowd spotters also watched for dense gatherings so officials could respond early, including by temporarily closing entrances.
The Fraunhofer Institute for Optronics, System Technologies and Image Exploitation is scientifically evaluating the plan. During the festival, it used GeoVID, an experimental system previously tested at other major events. The system uses AI to assess how many people are in an area and how crowds are moving.
The AI analyses crowds rather than individuals. It identifies where people appear in video without assessing faces or personal features. The footage is converted into heat maps and anonymised density grids, and the clear video is not stored.
The measures followed temporary overcrowding on a section of Wirtsbudenstraße in 2025. Munich plans to use the findings to improve safety arrangements for future Oktoberfests.
The University of the Bundeswehr Munich also used this year’s festival as a real-world research laboratory. Its work focuses on improving simulations of crowd movements, with the longer-term aim of forecasting visitor flows, identifying potential dangers early and modelling the effects of possible measures.