According to a recent whitepaper by PwC, only 9% of manufacturing companies have implemented AI processes to improve operational decision-making. That same study goes into detail about the reasons behind this slow adoption. Despite being hailed as a game changing technology, the business world is confronted with a complex, system wide challenge. Implementing artificial intelligence at scale is resource intensive and requires a deep level of technical expertise and strong leadership commitment. In follow up to my previous blog Understanding how AI ‘thinks’, this time we’ll take a look at 4 key ideas for mastering AI in manufacturing.
Understanding what AI is (and what it isn’t)
When we talk about AI, we do it in human terms. We talk about how AI “thinks”. We talk about vision and machine learning, neural networks, copilots…
Projecting human characteristics onto artificial intelligence leads us to have unrealistic expectations of what this technology is capable of. Artificial intelligence is a pattern finder, not a problem solver. Think about product design and innovation. AI can generate highly optimized designs based on pre-defined parameters. AI has zero ability to think outside of those parameters. Humans do. We call it “thinking outside the box” and it’s fundamental to creativity and innovation.
Generative design will always be incremental design – there may be millions of stages and iterations, but there will always be a traceable path from one design to another. AI, in its current state, will never have a “eureka” moment and develop something totally new. It can’t observe the way certain seeds stick to your clothes and invent Velcro, it can’t experience being uncomfortable in a chair and dream up a revolutionary new design. AI complements human capabilities, streamlining tasks and provides tools that support our own creativity.
Understand the application
To a man with a hammer, every problem is a nail. You can’t choose the right tool for a job unless you have a good understanding of the nature of the problem you’re trying to fix.
Build a thorough understanding of the context and root causes of the problem, and its implications. Then you can be more confident that the situation is suitable for an AI solution and if so, you can choose the most appropriate technology for the job, machine learning, language models, robotics, generative design, whatever it may be.
Creating this level of understanding hinges on collecting the right data, in terms of quality and quantity. High quality data will provide the basis for identifying and defining the problem. Within this data, an AI model will be able to find patterns, trends and correlations that might not be evident otherwise.
Once the problem is identified, detailed analysis can help dig deeper to understand the root cause and the contributing factors. Quality data ensures the accuracy of this analysis, while the quantity of data improves the reliability and representativeness of the analysis.
Understanding the changing technology
Technology and data infrastructure are the backbone for artificial intelligence applications. Yet this technology is rapidly evolving and that adds to the uncertainty of investment decisions. New tools, platforms and hardware are constantly introduced.
This changing business environment highlights the need for ongoing investment, training and research to keep abreast of the latest developments. Scalable infrastructure like cloud computing and flexible software licensing will help organisations adapt.
In addition to the technology requirements, there is a simultaneous need for the right people with the right skills and knowledge to install, operate and maintain this infrastructure. Data science roles are increasing in their breadth and diversity, and there is a need to invest in a continuous learning programme to keep these people up attract the best talent and keep those individuals operating at their peak.
Building the right strategy
Think strategically. This is a long-term project so it’s important to get a plan in place. First clarify your objectives and assess your current situation. Identify the business problems AI can help solve and evaluate your existing technology, data and skill levels to determine your readiness for AI projects.
At the beginning, it’s worth prioritising initiatives based on the value they are likely to generate and their ease of implementation. Choose the appropriate technologies to suit your needs. Depending on your current capabilities, you may need to bring new skills into the organisation. Consider whether to upskill existing staff, hire new talent or work with external agencies or consultants.
Building a great AI strategy is about more than just implementing technology. It involves a significant cultural shift. You’ll be adopting new tools, new processes, new staff and new ways of thinking. This change must be managed carefully and strategically but when you get it right, it has the potential to significantly increase efficiency, drive innovation, lead to improved decision-making and give your company a competitive edge in the marketplace.
And now we want to hear from you. What do you think are the keys to AI success in manufacturing? Share this blog with your networks and join the conversation online.
