The manufacturing industry is witnessing a transformative era with artificial intelligence (AI) integration. As Anna-Katrina Shedletsky, a former Apple engineer and an expert in the field, points out, AI may seem like a new buzzword. Still, its underlying technologies, such as computer vision, advanced statistics, and machine learning, have been around for decades. The key to harnessing AI’s full potential in manufacturing lies in understanding its applications and implementing it strategically.
The evolution of AI in manufacturing
In the realm of manufacturing, AI is not entirely new. Its applications span across various domains like product development, design, operations, and field quality. AI in manufacturing can be broadly classified into two categories: Component AI and Systemic AI. While Component AI focuses on solving specific problems in the manufacturing process, Systemic AI offers value across the system, such as optimizing production lines using multi-variant inputs. Although General AI, which deals with unstructured data to find insights, remains largely theoretical, the current focus is on leveraging Systemic AI for tangible outcomes.
Shedletsky’s experience at Instrumental, a company specializing in Systemic AI for product quality and efficiency, underscores the substantial benefits that AI can bring to manufacturing. By adopting AI as a global quality system, companies can proactively identify issues, accelerate training, and prevent errors, leading to significant cost savings and high returns on investment.
Four-step strategy for maximizing AI returns
1. Building a business case for AI integration
The first step in integrating AI into manufacturing is constructing a solid business case. This involves identifying at least three value drivers: reducing labor, increasing yield, or minimizing waste. A detailed calculation of the expected savings and improvements in these areas will form the backbone of a return on investment (ROI) case, justifying the costs of implementing AI.
2. Data aggregation in the cloud
The effectiveness of AI depends on the availability and accessibility of data. Manufacturing data, often siloed, needs to be centralized and structured for AI to be trained effectively. Shedletsky advises partnering with experts who can aggregate this data and extract value from it, highlighting the importance of cloud-based solutions for future-proofing and remote access.
3. Choosing the right AI technologies
Selecting AI technology should be guided by the business case. Technologies that require specialized AI engineers or data scientists might not be the best choice unless such resources are already available in-house. The focus should be on AI applications that are easy to train and align with the team’s existing expertise.
4. Establishing proof of value
Implementing AI technology should be approached as a Proof of Value (POV) rather than a Proof of Concept. This shift in perspective emphasizes the impact of the technology on the value drivers identified in the business case, ensuring that the technology not only works as intended but also delivers tangible benefits to the manufacturing process.
The future of systemic AI in manufacturing
The potential of Systemic AI in manufacturing is immense and can be realized relatively quickly with the right partners and a value-first approach. Shedletsky’s insights point to an exciting future where AI can significantly reduce the cost of quality in manufacturing, offering superhuman capabilities in identifying and solving problems. This era in manufacturing is not just about integrating new technology but about strategically leveraging AI to enhance efficiency, quality, and profitability.
Integrating AI into manufacturing is not just about adopting new technology; it’s about strategically applying AI to enhance efficiency, quality, and profitability. By following these four guidelines, companies can maximize the returns from AI investments and pave the way for a more efficient, cost-effective, and innovative manufacturing landscape.
