There is a shortage of artificial intelligence (AI) talent, with few industrial companies possessing sufficient AI personnel internally. AI will transform many jobs, and businesses should provide every employee with the necessary knowledge and training to adapt to their new AI-enhanced roles. AI resources can help businesses achieve new business models and provide better services.
Over the past decade, the design, development, and implementation of artificial intelligence have expanded into many fields. Manufacturing companies are working to understand the commercial potential of AI and to find suitable AI talent.
More and more countries are recognizing the opportunities presented by artificial intelligence and are beginning to develop national AI strategies. Finland launched its AI initiative in 2017, making it one of the first countries to do so. This AI initiative identified a small group of companies as pioneers in AI implementation; most companies are in the early stages of using data and AI in their operations.
Bridging the skills gap in AI
One way to address the AI skills gap is to increase resources for digital, mathematical, and technical education. In Finland, for example, the current education system does not adequately emphasize the application of AI across different fields. Academic and training programs are failing to keep pace with the rapid innovation in AI. AI education should begin as early as possible and be integrated into every stage. Officials from academia, companies, and the public sector must work together to ensure comprehensive AI courses are available. Massive Open Online Courses (MOOCs) offer a new avenue for providing basic AI education to the general public, which is a very good approach. However, deeper understanding often requires tailored educational modules.
Compared to many other industries, manufacturing is currently lagging behind in the application of AI and machine learning (ML). Adopting new technologies, especially in process industries, requires lengthy planning, which is time-consuming. Manufacturing companies have a long history of optimizing production, and their investment lifecycles can last for decades, making rapid changes difficult. Furthermore, safety and environmental regulations require stringent oversight.
According to PwC's AI Impact Index, by 2023, operating profit margins (the percentage of revenue remaining after deducting the cost of goods sold and operating expenses) in some sectors could increase by 60% to 100%. The "AI improvement curve" may differ across industries, primarily influenced by two factors: 1) the speed at which industries adopt different AI applications; and 2) the development of AI solutions that address specific industry problems.
Benefits and challenges of AI manufacturing
In manufacturing, short-term benefits are expected to primarily come from process automation and productivity-based solutions. Medium-term benefits will stem from the enormous potential of intelligent automation, enabling the automation of more complex processes, while predictive maintenance and optimization applications can further improve performance.
Dimecc's Machine Learning Academy showcases the cyclical nature of artificial intelligence and machine learning projects. Image credit: Dimecc
The productivity gains brought about by AI and ML depend not only on the introduction of the technology itself, but also on changing the way work is organized and expanding employees' knowledge.
Research indicates that the biggest obstacle to adopting AI and ML is the skills gap. Most of the time, surveys point to the technologies required to develop AI and ML solutions. However, the biggest skills gap in AI and ML is present across organizations.
The final report of Finland's AI initiative points out that while Finland provides high-quality education for those aspiring to become AI professionals (in information technology and mathematics), gaps exist in the application of AI. These are the areas where AI's effects will be most quickly apparent. The working group states that to achieve ambitious AI goals, it is crucial to ensure a diverse range of educational opportunities, invest in new educational approaches, and develop new talent attraction programs.
Continuing employee education is a challenge, and different operations and mechanisms can address these issues. A key factor is raising managers' awareness and understanding of AI opportunities to ensure sufficient investment in new, more flexible educational approaches.
Requirements for employees' AI skills
Employee skill requirements are influenced by changes in job market demands. The need for new talent in the development and application of AI is growing rapidly. Traditional educational pathways cannot meet this demand. New operational methods and mechanisms are needed to effectively enhance the AI skills of existing employees.
Generally, most of an employee's skills are acquired through on-the-job learning; therefore, companies bear greater responsibility for their employees' skill development. Companies should actively seek opportunities to educate and train their employees, either internally or in partnership with other organizations.
There are many educational methods, but few focus on on-site learning in an Industry 4.0 environment. Companies need appropriate performance evaluation strategies and employee training, as well as self-regulation, reflection, collaboration, and blended learning to mitigate the risk of excluding employees from the Industry 4.0 environment. Companies that lack proper training will experience impacts in productivity, product diversity, and quality.
Businesses need to equip existing professionals with AI skills in order to leverage their knowledge in AI-driven environments. A 2018 study on the “future work environment” and the “learning home” supports this argument, suggesting that training employees in AI and ML skills may be an effective way to bridge the skills gap.
The success of employee training will depend on their flexibility, problem-solving abilities, and willingness to engage in lifelong learning; otherwise, employees may struggle to keep pace with changes in the workplace and work processes. This challenge also explains why many companies are reluctant to invest in cyber-physical systems (CPS), which often include AI. Enterprise-level skills management and public education reforms are key factors in the adoption of CPS.
Machine learning courses and training
You can get free, general online training on AI and ML from major technology providers (such as IBM, Microsoft, Amazon, and Google) or MOOC courses organized by well-known universities.
For example, there's an online course called "Elements of AI," created in collaboration between the Finnish technology company Reaktor and the University of Helsinki. Typically, the purpose of such training is to "demystify AI," encouraging more people to understand what AI is, what it can do, and its limitations.
The Machine Learning Academy from Dimecc, in partnership with Futurice, focuses on industry-tailored solutions, utilizing targeted approaches to close, or at least narrow, the AI capability gap. The Academy's primary target audience includes R&D executives and engineers managing and involved in AI/ML development projects, as well as business and product owners. To accomplish these tasks, they need to understand how to define, plan, evaluate, and manage development or in-house projects that incorporate AI and ML elements. For example, for R&D engineers, it's important to understand how the introduction of these new technologies will change the functionality, boundaries, scheduling, and interfaces of their product development processes. Upon completion of the course, participants will have a foundation in AI and ML and the ability to identify and manage development tasks designed to benefit from these new approaches.
Throughout the course, various types of business and technology will be introduced and used as learning tools. Their main purpose is to help participants understand the key focuses and stakeholders they need to engage with at different stages of a data science project. For example, the "Business Objectives and Context" module in the first section guides users to work with business owners and those funding the project to answer questions such as "What are the business objectives of this project?" and "How can they align with our business strategy?"
Participants were excited to learn more about how ML projects drive and shape real-world business. Furthermore, topics related to preparing and running actual ML projects received attention, such as data preparation (collection, cleaning, preprocessing, filtering, analysis, etc.) and comparisons of different ML methodologies. One participant remarked, “In many cases, a lot of the work we do just confirms that we don’t have enough data.”
A concrete example is Ponsse PLC's field project, which focuses on after-sales service, particularly on-site maintenance of crop harvesting equipment. In this project, hydraulic oil (ML) was used to identify the required oil change intervals. Currently, the hydraulic oil and filters are changed every 1800 hours, while the optimized change intervals mean significant cost savings.
Skills gaps in machine learning applications and training have implications for the manufacturing and mechanical engineering industries. While these gaps will persist for the foreseeable future, there is a clear need for tailored AI/ML training programs to help companies train their employees and encourage them to start experimenting with AI.
Four suggestions for improving AI/ML training
To improve the effectiveness of AI training and enable more manufacturing companies to benefit from it, the following four suggestions are offered to companies, industry associations, and other organizations providing AI and ML training:
1. Customize courses for your industry
We advise against directly competing with leading tech companies like Google in the AI arena. Instead, we suggest becoming a leading AI company in your industry, where developing unique AI capabilities will give you a competitive edge. How AI impacts your company's strategy will depend on the industry, your company, and specific circumstances.
2. Focus on company-wide training.
We recommend against establishing a separate AI department within an organization. Instead, we suggest enhancing AI capabilities and understanding at all levels, from management to the workshop.
3. AI training should encourage specific pilot projects and use cases.
Developing AI training courses that encourage specific applications can help transform AI concepts into real-world value.
4. Improve existing AI education
Explore opportunities to establish dedicated AI education accounts to facilitate the smooth operation of the adult education market. Increase the number of web-based training courses and make university courses available to everyone. Integrate AI education with vocational school curricula.
Key concepts:
■ There is an AI skills gap in ML applications.
■ Industry 4.0 requires more AI and ML knowledge.
Think about it:
Attracting the next generation into manufacturing requires improvements to education and training mechanisms for ML and AI.
