Initial Situation

In industrial halls, gates are often opened completely, regardless of whether a person, a passenger car, or a large utility vehicle is passing through. This leaves an unnecessarily large opening for an unnecessarily long time, allowing heated or cooled air to escape. The goal was to find a control system that recognizes what is approaching the gate and automatically adjusts the opening time and height accordingly.

Eine Hand, die ein Smartphone hält, auf dem eine Energie-Rechner-App geöffnet ist; im Hintergrund sind Maschinen und Garagentore zu sehen.

Project Overview

Client
Bircher Smart Access

Area
Industrial & Finance

Solution
Adaptive Gate Opening using Stereo Image Processing.

System Environment
Stereo sensor with two synchronized CMOS cameras and fisheye optics, embedded Linux computer, Bluetooth Low Energy, Ethernet for connecting a second stereo sensor for both gate sides

Functions
3D measurement, classification, and tracking of objects, automatic calibration (plug and play)

Availability
Commercially available since 2019

SCS’s Contribution

Together with Bircher Smart Access, SCS developed Hyperion 3D as an integrated system comprising electro-optical hardware, embedded software, stereo image processing, and calibration software. The system measures, classifies, and tracks objects in three dimensions, thereby controlling adaptive gate opening.

Opened Only as High as Necessary

Based on an object’s size, speed, direction of movement, and type, Hyperion 3D determines when and how far the gate needs to be opened. This ensures it opens only to the required height for each vehicle or person.

Cross-Traffic Filtered Out

Objects merely passing by the gate do not trigger unnecessary openings. This reduces the number and duration of unnecessary gate openings.

Objects Detected in Real Time

The stereo vision algorithm dynamically learns the three-dimensional background and determines the size, speed, direction, and distance of approaching objects in real time. Machine learning methods differentiate between cars, forklifts, trucks, and people. Influences such as strong light exposure, shadows, rain, and snow are also taken into account.

Systematically Tested Under Real Conditions

A special test setup and extensive reference recordings enabled automated testing of hardware, software, sensor technology, and required gate behavior. This allowed the entire system to be reproducibly verified in various situations.

Wärmebild einer Person, die vor einem Lagerhaus mit Laderampen einen Gabelstapler fährt.

Result

Hyperion 3D resulted in a commercial sensor system that specifically reduces the number, height, and duration of industrial gate openings. It detects cross-traffic, differentiates between various road users, and adapts the opening to the respective situation. This reduces energy losses and associated costs. If required, two stereo sensors monitor both sides of the gate and self-calibrate within minutes during installation. SCS was responsible for significant parts of the development, including system design, electro-optical hardware, embedded software, image processing, and calibration software. Hyperion 3D has been commercially available since 2019.

since 2019

commercially available sensor system.

4 Object Classes

Car, forklift, truck, and person differentiated by machine learning.

Adaptive Opening

Height and timing dependent on the detected object.

Services Used

Computer Vision

Embedded Systems

Electronic Hardware

Industrial & Finance

A man in a blue shirt, a member of the SCS team, stands in front of a colorful yellow-blue background.

Your Contact Person

Michael Rusterholz

AI & Robotics
+41 43 456 16 00

For questions regarding stereo image processing, embedded systems, and real-time 3D object recognition.

More Case Studies

  • Ein Laptop, der an wissenschaftliche Geräte angeschlossen ist, steht auf einem Tisch in einer Laborumgebung.

    Georg Fischer

    Vision for Groundbreaking Weld Seam Inspection

    Image-based inspection of weld seams.

  • View through a car windshield with measurement points, parked cars, and a triangular traffic sign in the foreground.

    ZF

    SOC/FPGA: 4D Full-Range Radar

    Radar signal processing on SoC/FPGA, ASIL B compliant.

  • Close-up of a microchip and components on a green printed circuit board of an embedded system.

    When the Microprocessor Is Discontinued

    System-on-Module with Linux instead of 1:1 replacement.