The young company Swarm Analytics has found a way to install crowd monitoring systems using standard camera technology, clever algorithms and without significant IT hardware - i.e. at low cost. People can be targeted as moving motifs and their paths can be tracked and to a certain extent guided. What sounds like surveillance, however, is merely a method for detecting human movements in complex environments and in large groups - plus AI (artificial intelligence) drawing conclusions from it. Swarm Analytics' software can detect and analyze motion in real time, but not search or detect people. In any case, measures can be derived in a flash that make processes with many people involved work better. This helps, for example, in shopping centers, parking management, at traffic junctions and in public transport. The system from Swarm Analytics is excellently suited for corona prevention. Where distances shrink and the risk of infection arises because crowds of people are about to form, the algorithms recognize this and can sound the alarm in good time. The low technical and financial effort makes it additionally attractive for operators. Cameras are now attached to almost every street corner anyway. They could start their prevention work almost anywhere and overnight.
The first customers have already jumped on the bandwagon. The mountain railroads in Ischgl, for example, which wanted to use the system to count and steer skiers, can now use it to warn against standing too close together, always exactly where it is needed. The Salzburg Advent market near the cathedral is planning an operation - if it is allowed to open. The Innsbruck city administration already tested the system in summer and ordered 28 licenses. If the city's Christmas markets are allowed to open, Swarm Analytics software will be able to detect crowds of people as they arise. Kitzbühel and also St. Johann are considering the use of the system. Swarm Analytics now wants to increase from 13 to 20 employees in the short term. "The pandemic has accelerated business," says Alge. "Previously, purchasing decisions took forever, now it's suddenly happening very quickly." A great deal of importance is attached to anonymity: apart from the fact that faces are not recognized, the images retrieved from the cameras are deleted every 40 milliseconds. The start-up emerges from these months as the clear winner of the crisis.