Wearables Smart glove Optimizes Production Quality in Manual Assembly

Related Vendors

VOSS Automotive GmbH

dataTec AG

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

Wuxi InfiMotion Technology Co., Ltd.

Despite increasing automation in production, manual assembly is still required in many places. With an AI-supported wearable for assembly in the automotive industry, processes can be monitored, workers can be relieved and error rates can be minimized.

The number of plug connections in production is increasing, which is particularly evident in the automotive industry as a result of the trend towards modularization and electrification of products. Depending on customer preference, each variant of a car can have different configurations of infotainment systems, electric seat adjustment options and numerous other components. A key challenge is that visual barriers often hinder the work of the fitters. Added to this is the time pressure under which the individual connectors have to be connected. Assembly workers are therefore confronted with an uncomplicated, but in the long term monotonous task - a potential source of error that can have critical effects on people and machines in an emergency. Such inaccuracies can have a variety of consequences, ranging from simple malfunctions to serious safety risks. It is therefore essential to monitor assembly and embed strict quality controls into the process.

To remedy this and optimize processes, the supplier Voss Automotive has developed a smart wearable called Click ID, which recognizes correct mating processes and thus makes manual assembly easier and less prone to errors.

Combining the Quality of Technical Progress and Human Skill

In modern production, the automation of processes continues to increase. In principle, efficiency and speed can be increased and error rates reduced if machines now take over work steps that not so long ago had to be carried out entirely by human hands.

Nevertheless, there are still specific activities in the production process that require a particular degree of skill and cannot be left to automated assistants. These are often highly complex and require precise movements with the appropriate dexterity. Here, as in many other areas, people are involved who not only carry out fine machining operations, but can also react individually to unforeseen situations. A quality that machines usually still lack.

Despite their adaptability, however, humans are also a potential source of error - especially when fatigue or distraction occur during repetitive work steps. There is also the risk of human error.

The most logical step is therefore to bring together the special quality features and strengths of people and technology, made possible by Click ID by Voss.

A Handy Solution with Integrated AI

The smart wearable in the form of a glove was developed specifically for monitoring the assembly process based on years of expertise in connectors and system solutions as well as industry knowledge in the automotive sector. The individual components are simple in theory: a wearable tool is attached to a textile carrier. Fitters wear the wearable - either with or without a separate work glove - throughout their entire working day, without affecting the company's processes. A variety of signals are recorded during assembly using motion and audio sensors.

An AI-supported analysis of this data enables real-time evaluation and determines whether the respective assembly step has been carried out correctly. Click ID then uses a light and vibration signal to provide feedback to the wearer, who can react immediately to possible errors and correct them.

Simple and seamless integration is made possible by data communication between the production IT and the device. The acoustic and motor data is pre-processed and fed into an artificial intelligence system. The entire testing process is documented by storing the data on a local PC, a local cloud or the Voss Cloud. Here, the data can be analyzed and used to train the AI and continuously improve its performance. In addition, precise conclusions can be drawn about the performance in each application in order to increase quality and efficiency in assembly.

For the functionality of artificial intelligence of this kind, the determined data signals are converted directly into so-called features. Based on this, a model is trained that learns typical feature values for a use case. This AI model is executed on the device and further developed and optimized through repeated retraining.

The smart glove can also switch between different use cases - an important feature that allows flexibility and quick adaptation to different tasks and situations. For workers who, for example, change jobs by the hour - from plugging in cables to assembling electronic components, for example - the model can be quickly replaced via device management.

Introduction in Five Simple Steps

From the initial definition of requirements to successful implementation in the production process, Voss has developed a five-stage process that is individually specified for each application.

• Getting to know each other: The first step is to identify potential applications for the glove. This involves consultation with both internal experts and customers to identify specific requirements and needs.

• Specification: In the following, the technical requirements are recorded and initial measurements are carried out in the production line, which serve as the basis for the proof of concept. This allows the added value and feasibility of using Click ID to be designed.

• Training: The integrated AI must be trained to critical performance. As much data as possible is collected and analyzed in order to optimize algorithms and generate more precise results.

• Installation: The AI and Click ID software are configured and commissioned individually. The devices are installed and tested in the production area.

• Support: Long-term maintenance and support are necessary to carry out regular updates, clarify questions and demonstrate application changes.

If these steps are implemented correctly, Click ID offers a new way of monitoring manual assembly, supporting workers with the help of AI and thus minimizing errors in production. The device is currently in the pilot phase and is being tested by the first interested parties. Additional certified products will be available from the third quarter of 2024.

*Dr. Anne Völkel is Head of Innovation Projects at Voss Automotive.

Machine Authentication

Machine Operator Authentication - Why knowing "who" matters