By: Manabendra Mukherjee, Project Manager, Digital Consulting & Advisory, LTI

The Manufacturing Industry has traditionally been labor-intensive and relatively slow in adopting automation. However, if any manufacturing process has been at the cross eye for automation for past 30 years, it would be Welding. Robot-supported Automotive Welding has transformed the automotive assembly lines in terms of safety, cost-efficiency, quality and process efficiency.

In the past, premium quality, durability, high levels of precision, fit-and-finish and structural integrity were associated with the luxury segment of vehicles. Today, they are expected of the mass-produced models as well. To achieve these quality goals in high volume manufacturing, global Auto manufacturers have invested in digitally-controlled automated systems.

Welding is a key manufacturing function, providing cost-effective and high-volume parts. It captures and tracks a wide range of production data. However, it comes with one limitation – it has traditionally required manual expertise. Engineers periodically test the weld assemblies and adjust the weld parameters.

Welding inconsistencies

Welds may be inconsistent due to various reasons – there could be misaligned electrodes, coating material composition or thickness, sealers, weld force, shunting and machine tooling degradation, etc. “The Wear of Weld Gun Tip” is also another factor that impacts weld reliability. Weld tips are dressed with cutters after a fixed number of welds in order to remove material deposits and improve resistance parameters. However, these weld tip dressers tend to lose sharpness over time. Also, expulsions due to spark increase the risk of injury and wastage of weld material.

Adapting to automate and predict

Enter Adaptive control welding systems – like those made by Bosch Rexroth – use unique hardware and software to adjust welding parameters during real-time production. They adhere to stringent tolerance limits, and monitor parameters like current, voltage and resistance every millisecond! Such a system adjusts the weld time and current applied during welding, adhering closely to master curve parameters, to produce quality welds.

Advanced Analytics also plays a critical role in today’s automotive manufacturing. Many leading auto manufacturers are collecting volumes of data from the vehicles, production process, equipment, supply chain, as well as structured and unstructured consumer/social media feedback. Leveraging the data provides insights as varied as how to improve the manufacturing quality by reducing detection to correction cycle time, how to make accurate demand forecast, how to optimize production time by predicting Equipment health, how to predict vehicle usage and breakdown by analyzing engine and sensor data, etc.

Analytics can play a crucial role in identifying weld defects and predict conditions before they get worse.

Addressing shortcomings

However, even with Adaptive Welding, manufacturers cannot gain real-time visibility into every instance of weld across hundreds of Robo Welding taking place, let alone identify potential areas where the quality is suspect.

Some ways Advanced Analytics is solving weld quality related issues like continuous monitoring of key weld parameters and proactive countermeasure, leveraging historic data to design future strategies, and predicting upcoming results based on analytical models. Together, these methods help enhance the production quality, train analytical models and estimate accuracy of model output, and calculate the probability of upcoming inferior welds to within an acceptable level of accuracy.

Soldering the future

With the advent of Industry 4.0 and associated deluge of data from connected vehicles, sensors, equipment, supply chain, consumer and external sources, and social media feeds, we now witness the rise of Data- as-an-Organizational-asset and Data-as-a-service. Moreover, new-age disruptions, along with Open Source technologies and cloud platforms will continue to redefine the automotive sector in the future.