It’s not an easy time to be in the automotive manufacturing business. A global pandemic has upended supply chains and wreaked havoc on auto sales in general. In parallel, automotive factories are increasingly connected through Industry 4.0 – the melding of the physical and digital worlds – and solutions such as robotics, remote sensors and digital control systems, which produce seemingly unending (and overwhelming) new streams of data to be processed and interpreted.
But – pandemic notwithstanding – perhaps no challenge is more pressing than the skilled labour shortage in the Canadian automotive sector. As the baby boomer workforce retires, it’s getting more and more difficult for manufacturers to find experienced workers with the technical craftsmanship, which is difficult to teach and transfer, to replace their predecessors and build vehicles with increasingly integrated computer technology.
Furthermore, the ability to share knowledge across skill and experience levels to solve problems becomes more difficult as turnover increases. Canadian automotive manufacturing companies are in a race to find solutions that enhance the capabilities of newer, less experienced workers for increased efficiency in training and onboarding. Canada must ensure that the next generation of skilled automotive workers have the support and knowledge required to take the industry into the new millennium, maintaining its reputation as an automotive producer known for quality and precision.
Here are three ways that AI is being applied to transform the Canadian automotive sector to help mitigate the effects of the diminishing skilled labour force seen across the industry.
AI can give employees the information they need, quickly and in context
When new workers are starting in their roles, they will need to search for a large amount of information critical to doing their jobs correctly and efficiently. But it isn’t just new employees that need to learn and access previous experiences. For example, quality analysts might search for information related to previous incidents similar to the one they are trying to solve, or maintenance technicians may search for knowledge on a particularly tough repair case. This data is often scattered across various systems and formats, including documents and spreadsheets. AI can help employees access this information, and make sense of it, faster.
AI-powered knowledge management tools go beyond traditional “filter and search” capabilities. They ingest data from a wide range of different documents and database records, giving employees access to a massive amount of information. When the system is queried, it returns search results, sourced from a variety of systems, that are contextually relevant, allowing quality analysts and compliance specialists – both senior and junior alike – to accelerate the resolution of t cases. As analysts use the system, they provide feedback about which recommendations were most effective, and AI uses this information to learn and improve over time, capturing institutional knowledge that is at risk of being lost.
AI can separate important signals from noise
The proliferation of data, even when it is well organized and accurate, presents a tremendous challenge to even the most highly skilled worker, whether their function is in automotive supply chain planning, production, maintenance or finance. Identifying meaningful patterns in these mountains of data is critical to gaining a competitive advantage, but it is exceedingly difficult in today’s data rich environment.
AI models create the opportunity to analyze historical and current data streams to uncover never before seen patterns beyond most human’s ability. For example, how can one better understand the impact of incentives on the future production schedule? For one automotive client, an AI-enabled forecasting solution improved predictive performance of automotive sales over a three-months forecast horizon by five per cent, and further allowed a better understanding of the impact of incentives. This type of insight drives tremendous competitive advantage across the supply chain.
Likewise, AI models can be connected to IoT data streams in the factory to predict failures of major assets, prevent production downtime and reduce maintenance costs.
With AI uncovering the most important data signals and making informed predictions based on that data, the skilled worker can be free to execute on the most impactful strategies rather than spending time trying to understand the data.
AI can automate more complex inspection tasks
Anomaly detection is also an area where AI and computer vision have the potential to add significant value. Visual inspection and anomaly detection processes often require a high level of experience and craftsmanship when performed by a human. Visual inspection processes to verify quality in areas like paint, welding, and assembly are quite nuanced and the nature of these repetitive tasks—coupled with the high level of required skill – leads to challenges in staffing these positions.
Automotive manufacturers have increasingly turned to automated visual inspection systems to complement or replace manual visual inspection. However, these systems have fallen short where the inspection is more subjective and requires significant human experience for correct interpretation of quality. Now, advances in computer vision, and the AI techniques to explain the judgment of the underlying models, are driving AI inspection solutions that can be confidently adopted in a factory setting. This development has the potential to more broadly reduce dependence on humans for repetitive, manual inspection processes, freeing up quality departments to focus more on root cause analysis and continuous improvement.
Not all automotive manufacturers are ready for AI
While the opportunities AI affords are exciting and tangible, organizing the entire business to effectively tackle this issue is a major challenge. Deploying AI solutions is very difficult, and only one in ten organizations in North America is ready to implement AI solutions. But that’s the beauty of AI—it’s possible to start small and lean, with a specific use case, and scale from there.
The shortfall of experienced workers in automotive manufacturing isn’t going away. Businesses that succeed in integrating AI and ML support solutions will be better positioned to bridge the skills gap, optimizing production responsiveness and creating significant business value.
Dan Wilson is the head of sales for Element AI in North America, working alongside various industries to deliver high impact, AI-enabled solutions for retail, supply chain management, manufacturing, insurance and capital markets, with the goal to solve challenging business