On the Way to Perfection: How TRONTEQ’s AI-powered Visual Quality Control Improves ROQSTAR Ethernet Switch Production
31 Mar, 2023

In the world of production, technology has played a crucial role in ensuring the smooth running of operations. For TRONTEQ, a company specializing in the production of Ethernet switches for public transport, every detail counts. With tens of thousands of ROQSTAR Ethernet switches installed worldwide, TRONTEQ’s recent visual quality control approach is critical to ensuring the highest production quality of the devices.

TRONTEQ Ethernet Switch Production Visual Quality Control

ROQSTAR Ethernet Switches: The Go-To Solution for Public Transport

In the last few years over 40.000 ROQSTAR Ethernet switches have left the TRONTEQ production to become a best friend for so many vehicle devices in public transport such as displays, passenger counters, video recorders, cameras and routers. Each time our switches had to comply with different and sometimes challenging requirements.

From Ethernet Switches Connecting Cameras to Cameras Monitoring the Production of Ethernet Switches: The Role of Visual Quality Control at TRONTEQ

ROQSTAR Ethernet switches have been used in a wide range of applications in the public transport sector and beyond: From safety cameras to mission-critical multicamera systems on industrial robotic arms. We have records of applications, where our devices run together with 92 Gigabit Ethernet Cameras, delivering always the last bit and byte of information in sensitive industrial real-time control systems.

Now, it’s time for the cameras to pay back.TRONTEQ installs multiple high-resolution 2D cameras and 3D sensors at various production stages of the ROQSTAR Ethernet switches. Starting from the initial Surface-Mounted Device (SMD) assembly, to the surface inspection of the housings, and from the incoming goods control to the palletizing of the packets in logistics, a plethora of cameras capture high-resolution images, generating the input for image processing algorithms. The features extracted are further processed, and the final outcome is data – Big Data.

The all-seeing eye: A camera installed in a screwdriving robot for Ethernet switch production

The main objectives of TRONTEQ’s newly implemented visual quality control are traceability, quality control/quality assurance, and production management.


Traceability in production refers to the ability to track the movement of materials or products through the supply chain, from raw materials to finished goods. This involves documenting information such as the origin, composition, and processing of materials, as well as the identity of suppliers, manufacturers, and distributors involved in the production process. Traceability is important for quality control, compliance with regulations, and identifying and addressing potential issues or risks in the supply chain.

Quality control & quality assurance

Quality control and quality assurance are two interrelated processes in production that are used to ensure product quality. Quality control involves inspecting and testing products at various stages of production to identify defects and ensure they meet quality standards. Quality assurance involves establishing and maintaining processes and procedures to prevent defects from occurring in the first place, as well as ensuring that quality control processes are effective. Both processes are important for ensuring that products meet customer expectations and comply with relevant regulations and standards. 

Production management

Production management is the process of planning, organizing, directing, and controlling the production process to ensure that goods are produced efficiently, on time, and to the required quality standards. This involves managing resources such as labor, materials, and equipment to maximize productivity and minimize waste. The goal of production management is to ensure that products are produced in the most cost-effective manner while meeting customer demand and quality requirements. 

The Hot Part: TRONTEQ Approaches Visual Quality Control by Combining Traditional Image Processing with Deep Learning Techniques

At TRONTEQ, we take visual quality control seriously. Our approach involves utilization of both traditional image processing and deep learning techniques to ensure the production of high-quality Ethernet switches for public transport.

Traditional image processing is formed using pre-defined algorithms and mathematical operations to analyze and process images. We use this approach to perform tasks such as filtering, segmentation, and feature extraction.

Visual analysis of a ROQSTAR housing: 16 ports found

On the other hand, deep learning is a subset of machine learning that involves training artificial neural networks to recognize patterns in data. Deep learning algorithms can automatically extract features from images without the need for manual feature engineering, making it particularly useful for object recognition, image classification, and segmentation.

Detection of UV reflection on a coated PCBA

While deep learning has gained popularity in recent years due to its ability to outperform traditional methods in some applications, it requires large amounts of training data and computational resources. In some cases, we use traditional image processing algorithms to preprocess images before feeding them into a deep learning model, or to post-process the results obtained from a deep learning model. This approach allows us to leverage the strengths of each technique and achieve the highest possible accuracy.

“Congratulations to our engineers Nicola Helble, Patrick Schmid for starting completely from scratch and successfully integrating electrical engineering with software engineering, robotics with vision, and bringing it all to life in a nick of time! We are pretty sure that we are the only Ethernet switch provider in the Public Transport industry that pays such close attention to detail and quality. And what’s most important – our customers appreciate our commitment to excellence, and we’re thrilled to deliver on their needs.” – Konstantinos Theiakos, Manager of Operations at TRONTEQ GmbH, on the newly implemented visual control system.

More Stories Like This:

Coming Soon: ROQSTAR Label Manager

Coming Soon: ROQSTAR Label Manager

A new addition to the ROQSTAR Universe is on its way: ROQSTAR Label Manager is TRONTEQ’s new tool to facilitate the maintenance of ROQSTAR Ethernet switches. ROQSTAR Ethernet switches are made solely for use in public transport vehicles. Such being...

read more

Our products are fundamental for the digitalization in public transport. ROQSTAR M12 Ethernet Switches provide the network infrastructure for e-ticketing, passenger counting systems (PCS), dynamic passenger information (DPI) and closed-circuit television (CCTV).


Hoelzlestr. 3
72768 Reutlingen



Connect with us!

© 2023 | TRONTEQ GmbH