Manufacturers today perform a delicate balancing act. As they deal with rapid changes in markets, fluctuating raw material and production costs and a shortage of skilled workers, they must also address regulatory pressures and rapidly evolving customer demand and requirements. Many manufacturing leaders are looking to the power of artificial intelligence (AI) to help them adapt to these pressures and become more innovative, efficient and competitive.
The 2023 Voice of Our Clients (VOC) research points to growing evidence that manufacturers are taking small but decided steps toward adopting AI. The percentage of manufacturers implementing AI initiatives rose four percentage points from 2022 to reach 18% in 2023.
The industries that are furthest along in the use of AI are mainly within the financial services and retail sectors. This is not surprising as they have a strong customer focus, are often transaction-heavy and must contend with a lot of competition. They have understood that staying ahead requires using AI to be more innovative and efficient.
In manufacturing, however, there are significant differences in AI maturity between organizations and even departments within the same company for many reasons. While many manufacturing organizations are traditional, some are entirely greenfield and do not have the legacy burden; some are more innovative in their approach, while others operate with a traditional culture. Additionally, in the past, manufacturers have not faced the same pressure from customers to digitally transform like service-based industries.
How can AI be applied in manufacturing?
AI promises many benefits for manufacturers across operations—from troubleshooting, preventive maintenance, streamlining, logistics, tracking purchase prices and delivery quality to marketing, customer experience improvements and personalization of offers and product design.
As my colleague Diane Gutiw notes in her blog Embracing responsible AI in the move from automation to creation, AI is neither new, nor is it magic. However, responsible AI can be an engine for efficient operations and offer new solutions to manufacturers' challenges. The best approach is to start with a problem the business wants to solve and evaluate which technology can provide the best results: more traditional automation or different types of AI and machine learning.
A good manufacturing use case is in product development, particularly product design. AI can quickly explore and evolve product innovations. It can also support the test phase of product design using data to improve product functions based on customer user data. Recently, Google’s DeepMind AI tool, GNoME (Graph Networks for Materials Exploration), helped predict the structure of millions of new materials that can be used in battery production. It can also support the cost-effectiveness of production by offering suggestions on maximizing yield and doing more with less.
With these empowerments, manufacturers are much better placed to advance their sustainability goals. In fact, according to the VOC research, 82% of manufacturers cite sustainability as highly core to creating future value. As GNoME illustrates, AI can empower manufacturers to identify and understand the most sustainable products and materials, improve production efficiency and use less energy.
Becoming more service-orientated with AI
The 2023 VOC results also indicate that industry and business priorities are pivoting toward creating more value by offering services. Across sectors, as marketplaces change and new ecosystems emerge, manufacturers seek to develop and deliver new services to serve their customers better and generate growth.
By collecting more data and information about customers' needs, usage and support requirements, and by applying AI to analyze this data, manufacturers can uncover entirely new service and revenue opportunities while improving the customer experience. For example, applying AI to data from “smart” products will enable manufacturers to offer value-added services that could lower the need for maintenance or provide suggestions on how the product can be used better.
There are several ways manufacturers can use AI to broaden their services—from frictionless aftersales services and tailored upgrades‑as-a-service to personalized insights and recommendations. For example, many manufacturing companies are increasing their turnover thanks to aftermarkets such as spare parts, support and warranty. The everything-as-a-service (XaaS) business model is becoming more prevalent, where the servicification of products leads to increased customer experience, helping to create deeper, long-term relationships with the customer. AI can support this evolution by automating and optimizing services and helping better understand customers – who they are, what they like and how to delight them.
Taking an agile and inclusive approach to AI
Building a business case for everything related to AI is impossible. Businesses must dare to experiment to get quick indications of whether a project can deliver value. This requires embracing a more agile way of working and welcoming the best ideas and use cases from everyone in the organization.
For example, a German rail company had to take their trains out of service for maintenance whenever the wheels of the trains began wearing out. It turned out that the train drivers with a lot of experience could tell just by the sound when the wheels needed replacement. We worked with the company to mount microphones on the trains, use AI to analyze the sounds and provide early warnings when it was time to change wheels, helping the company save costs, provide better safety, and benefit travelers. Notably, this AI solution was not based on pre-existing data.
This change in culture and openness to innovation can challenge many traditional manufacturers. At CGI, we firmly believe that a human-centered approach to transformation is critical for success.
Learn more about how we combine responsible AI with human creativity to deliver pragmatic and innovative projects, or contact our manufacturing or AI experts.