Customised computer vision for lifts and escalators

Computer vision makes it possible for machines to analyse data and extract relevant information from them. Based on this, they then send signals to controllers, displays or other devices.

Elfin Technology GmbH has developed a computer vision solution in the form of Elfin Vision that is especially adapted to the requirements of lift operation.

Computer vision can markedly improve the safety, user-friendliness and efficiency of lifts. This technology is a subdivision of artificial intelligence which aims to extract information from visual data. The source can be photos, videos or multidimensional data, such as time-of-flight sensors. Computer vision mimics the human capacity for recording, processing and analysing pictures. The algorithms are already superior to people in several areas.

Computer vision is based on three building blocks: picture processing, pattern recognition and machine learning. During picture processing, visual data are detected and categorised. Machines are able in this way to detect patterns. Computer algorithms then use these patterns to make predictions and decisions.

Computer vision in lifts

To enable computer vision to provide reliable signals in lifts and on escalators, the sensors must detect patterns correctly under the most difficult conditions: during poor and changing light conditions, in cramped spaces, from various angles, in reflections and in the presence of large, constantly changing collections of people and objects.

Only if this succeeds can computer vision create real added value in lifts. Elfin Technology GmbH has now launched a computer vision solution in the form of Elfin Vision that is especially tailored to the requirements of lift operation.

Detect dangers and avoid accidents

Photo: © Elfin

According to the manufacturer, Elfin Vision sensors are capable of reliably identifying people, movements and objects even under the challenging conditions in lifts. For example, they detect fast-moving people approaching the lift doors or people with prams who want to step on to the escalator. They then send signals that trigger visual and acoustic warning sounds or stop the controller of the doors and escalators.

In addition, they can register collections of people at critical points (such as the end of an escalator), report this to the controller and in this way avoid dangerous situations. Another use case for the Elfin Vision sensors are emergencies that have already occurred: for example, they provide information on whether and how many people are in the lifts and can detect whether they are lying on the ground unconscious.

Sensors are GDPR-compliant

Elfin Vision:
1. the sensors has a comprehensive basic training in picture and pattern detection in conveyance technology.
2. computer vision experts decide in each case what camera position, lens and resolution are ideal for the particular circumstances at the point of use.
3. the sensors undergo additional customisation training under real conditions. this is necessary to adjust the basic competences of the machine to the particular circumstances.

The Elfin Vision sensors operate autonomously, are GDPR-compliant and thanks to the open interfaces offer seamless integration in existing systems, the company emphasised.

More informations: Elfin Technology is a provider of smart solutions for lifts based in Cologne. Apart from Elfin Vision, the company develops smart multimedia displays and monitoring gateways for lift maintenance.

elfin.de/en