It’s not easy leading a manufacturing company these days. Executives are navigating a landscape fraught with economic constraints, intricate supply chain challenges, rising environmental and compliance requirements, and shifting customer expectations. Amid these pressing issues, Generative Artificial Intelligence (Gen AI) is emerging as a powerful tool to enhance manufacturing capabilities in areas such as predictive maintenance, process optimization and intelligent automation.
Significantly, Gen AI’s potential is being validated across the entire spectrum of manufacturing activities, from R&D to customer support. Consequently, the market for Gen AI in manufacturing is anticipated to reach approximately USD 7 Billion by 2032, with a robust compounded annual growth rate of over 40 percent.1 Beyond boosting operational performance, Gen AI is pivotal in fostering a more agile and competitive manufacturing environment.
In this article, we delve into a range of real-world applications of Gen AI in manufacturing, showcasing the groundbreaking potential of this technology.
Gen AI across the Manufacturing Value Chain
Forward-looking manufacturing companies have already begun their Gen AI journey. These early adopters are using the technology to re-shape decision-making processes, enhance production lines, elevate product quality and curb waste.
While approaches vary, manufacturers face the challenge of shaping their offerings, producing output and getting the product to market. In this paper, we’ve distilled a typical end-to-end manufacturing process into five critical stages. This serves as the backdrop for identifying and exploring Gen AI opportunities.
Case Study
Co-creating an Improved R&D System Through AI-led Data Processing
Read More
Case Study
Machine Learning Solution Prevents Duplicate Transactions
Read More
Figure 1: Gen AI across the Manufacturing Value Chain
1. R&D
Gen AI has a pivotal role to play right at the start of the manufacturing lifecycle. Its ability to rapidly create rich, detailed and technical content – such as innovative designs, models and simulations – turbocharges the creative process. Gen AI can suggest, test and analyze alternative product designs, layouts or specifications that can be scrutinized for performance, cost and aesthetics.
The scope of Gen AI can range from creative prototyping to enabling high-precision digital twins that test and refine late-stage product designs. It can also contribute to highly technical areas such as topology optimization, predictive design modeling and stress testing. Along with other types of automation, including AI and Computer-aided Design (CAD) software, Gen AI is an integral part of the emerging discipline of ‘Generative Design.’
Case in point: Airbus2 has accelerated its design processes by leveraging Gen AI, creating lighter, stronger aircraft parts. The technology creates numerous alternative designs based on pre-defined parameters, empowering the manufacturer to fine-tune options, accelerate prototyping and enhance collaboration throughout the design lifecycle.
Case in point: Siemens Gamesa3 is using digital twins to create new designs for wind turbines. The technology involves training Gen AI on large datasets about existing wind turbines. The system has been used to generate new designs that are materially more efficient and cost-effective than existing specifications.
2. Sourcing and Procurement
Obtaining raw materials, energy, components and intellectual property is labor- and data-intensive. Gen AI can support this process at multiple stages.
For example, during supplier selection, Gen AI can analyze and summarize data to identify potential suppliers and evaluate fit-against-tender specifications. In supplier risk management, Gen AI models can analyze extensive datasets for existing supplier performance, drawing on financial records, industry news and a range of other internal and external sources to model and predict the risk of supplier failure – for example, bankruptcy or supply delays. This enables businesses to better monitor and proactively address risks. Appropriate strategies and contingency plans can then be implemented to switch or diversify supplier base and mitigate any potential impact on operations.
Gen AI also enables better buying. The integration of Gen AI with historical sales, market trends and social media information enables manufacturers to more accurately anticipate demand fluctuations and purchase accordingly. This reduces waste, promotes sustainability and improves cash flow.
There are multiple other use cases. During contract negotiations, for instance, Gen AI can accelerate the analysis of contract terms, as well as monitor contractual databases for hidden discounts and imminent penalties. Gen AI can also automate and personalize communication between manufacturers and suppliers, such as the handling of queries, translation and high-volume / low-value price negotiations.
Case in point: Procter & Gamble4 uses Gen AI to enhance the accuracy of demand forecasting. By analyzing historical data, external factors, such as weather conditions, and prevailing market trends, the company aims to precisely predict consumer demand. This improved forecasting enables Procter & Gamble to optimize inventory management and reduce operational costs.
Case in point: A prominent US retailer5 sourcing products from various manufacturers employed Gen AI to build chatbots to expedite negotiations on costs and purchasing terms with vendors. This initiative, in its initial stages, has already demonstrated cost reduction by introducing a structure to effectively manage intricate tender processes. Moreover, two-thirds of vendors reported that they preferred interacting with the bot than a human.
3. Production
Gen AI is being used across the core areas of the manufacturing process – production, assembly and testing. For instance, Gen AI is used to enhance predictive maintenance in the production process. Proactively identifying and resolving potential equipment failures before they occur is a massive performance improvement opportunity – keeping production lines open and repair costs down. Gen AI can plug into the predictive maintenance process in different ways, such as in the analysis of large volumes of multimodal data (e.g., text, pictures, video and code), in modeling and forecasting performance during peak periods, or in the optimization of maintenance schedules to reduce breakdown risk. Recent analysis shows that predictive maintenance boosts productivity by 25 percent, slashes breakdowns by 70 percent and cuts maintenance expenses by 25 percent.6 Better maintenance also means longer asset lifetimes and replacement cycles.
Similarly, Gen AI is being integrated into the quality control processes of various manufacturers. Gen AI models can be trained to analyze images or videos of components and assembled items as they travel through the production process. After training with a dataset of defective and non-defective items, the model can work in real-time to identify subtle defects or deviations from the standard, enabling real-time intervention or replacement.
Case in point: A leading manufacturer7 employs Gen AI to refine and improve its production planning processes. The tool analyzes vast historical and real-time datasets (including volatile variables such as weather) and identifies trends and patterns against which better decisions can be made about the timing, location and volume of productions runs.
Case in point: ABB is collaborating with Microsoft8 to integrate Gen AI capabilities into its industrial applications for quality control and improved sustainability in operations. ABB is set to enhance its Genix platform and applications by integrating Gen AI through Azure OpenAI Service, including advanced large language models like GPT-4. This integration will introduce capabilities such as generating code, images and text. The forthcoming application, Genix Copilot, will improve user experiences by offering intuitive functionalities and facilitating the streamlined flow of contextualized data throughout various processes and operations.
4. Distribution and Logistics
Getting the product to market, either to retailers or direct to consumer, requires strong and flexible logistical capabilities. Gen AI has a part to play in meeting this challenge, not least in terms of inventory management. Gen AI can model future product demand by analyzing trends, historical data and external factors such as market conditions and seasonal influences. This can feed into improved inventory planning, reducing overstock and understock situations and optimizing holding costs.
In warehousing, Gen AI can optimize the layout and operations by determining the best ways to store goods, pick orders and replenish stock. In supply chain planning, Gen AI can simulate various supply chain scenarios to identify potential bottlenecks or disruptions.
In delivery planning, Gen AI can analyze traffic, fleet performance data, delivery deadlines and generate optimized delivery routes. This optimizes the number of deliveries per day and reduces wasted fuel and labor costs. The same applies to optimizing freight consolidation and utilization strategies.
Gen AI can also be part of the efforts to integrate multiple data types used across the logistics process. Think location, temperature, humidity, handling conditions and contaminants – Gen AI can analyze, summarize and make recommendations from different types of data, enabling companies to make informed decisions and minimize environmental impact.
Case in Point: One of the largest logistics companies in the US9 is harnessing Gen AI to enhance the efficiency of picking routes within its warehouses, driving a 30 percent increase in workforce productivity and significant reductions in operational costs through optimized space and material handling. Gen AI has introduced a new layer of customization, optimizing routes to use less fuel and prioritizing specific deliveries. This platform has allowed the company to evaluate and refine its trade network's efficiency, even providing actionable suggestions for further enhancements.
5. Customer Service and Maintenance
Gen AI is re-defining how manufacturers interact with customers, automating and personalizing a range of services from query handling to maintenance scheduling.
Gen AI-powered chatbots are transforming customer support in the manufacturing industry, streamlining and automating a spectrum of customer interactions, service tickets and complaints. These virtual assistants can efficiently manage issues such as product recalls or delivery problems, stepping in as alternatives to human representatives to address concerns and offer guidance, including troubleshooting steps. Where required, the chatbots seamlessly transition the conversation to a human agent.
Gen AI chatbots can be pivotal in keeping stakeholders (e.g. executives, engineers and planners) updated on delivery progress, offering real-time helpdesk and notification functionality covering status, timing and condition. This reduces inbound volumes, optimizes the efficiency of customer support teams and enhances overall customer satisfaction.
Gen AI is also improving condition monitoring and predictive maintenance for customer equipment. By interpreting telemetry from equipment, Gen AI empowers manufacturers to better diagnose anomalous performance and intervene to minimize the risk of breakages, where service revenues and warranty costs are important – in sectors like aerospace, automotive and white goods manufacturing. In the event of an issue, Gen AI offers recommendations and service plans to assist maintenance teams in resolving challenges swiftly.
Case in point: Siemens10 is integrating Gen AI into its B2B predictive maintenance solution. The upgrade is expected to make human-machine interactions conversational and intuitive, with interactive dialogues for streamlined decision-making for faster, more efficient predictive maintenance.
Case in point: GE Appliances11 has developed a Gen AI-powered app to create recipes, reduce food waste and make home cooking easier, thus fostering rich interactions with customers for the lifetime of their products long after purchase.
Case in point: Boeing12 employs AI-driven simulations to craft lifelike training scenarios for its pilots and flight crews, enabling them to hone their emergency response skills in a secure and monitored setting. Through the replication of diverse situations, AI contributes to bolstering crew readiness and sharpening decision-making abilities, thereby mitigating the risk of accidents stemming from human error.
The Gen AI Adoption Challenge
As this paper demonstrates, the manufacturing lifecycle is studded with Gen AI use cases, and the list is expanding rapidly. However, many Gen AI deployments are still in the experimental or pilot stage. Recent surveys 13,14 reveal that while 65 percent of US executives believe Gen AI will have a significant or highly significant impact on their organization in the next three to five years, 60 percent admit they are still a year or two away from implementing their first major Gen AI solution. This disparity highlights a notable gap between the potential and readiness of organizations to implement Gen AI.
Navigating the path to Gen AI might appear daunting, and leaders are looking for external help in areas such as strategy, prioritization, infrastructure requirements, algorithm development and optimization. We define success here as the ability to customize models, scale, adapt to diverse tasks, accelerate development cycles and ultimately enhance the overall effectiveness of AI solutions.
Gen AI in Manufacturing: The Horizontal Dimension
Like other sectors, manufacturing offers a range of industry-specific applications for Gen AI, as already demonstrated in this paper. Beyond these ‘vertical’ applications, manufacturers are deploying Gen AI across horizontal functions, including Finance, Procurement, HR, Sales and Marketing, and IT. So, leaders are not only deriving value from Gen AI in activities unique to manufacturing but also leveraging its benefits for common tasks across various industries. These tasks include hiring talent, managing supplier payments and providing IT support in both back-office and front-office operations.
Ramping up Gen AI: The Pathway to Success
Like other waves of technology-driven change in the manufacturing sector, Gen AI does not exist in a vacuum. Guiding and expanding its deployment in a manufacturing context demands management, expertise and effort. WNS recommends a four-fold approach for all manufacturers beginning their Gen AI journey:
While there are dozens of use cases across manufacturing, their complexity and payback will vary significantly for different companies. Run a structured process to identify the initiatives in terms of cost / benefit for the first pilots. Selecting the wrong project could stall momentum and lead to failure.
3. Access the right capabilities
Gen AI success requires multi-disciplinary expertise, encompassing business analysis, domain expertise, AI modeling / programming / optimization, systems integration, data science and analytics, communications and project management. Where these skills exist in-house, nurture them. Where there is a major gap or need for speed, understand the partnership needs to assemble the right skills. In manufacturing, as for other areas of the economy, most organizations are learning that you can’t move quickly in AI by doing everything home-baked.
Gen AI is a data-intensive technology, extracting new value from data in its many and varied forms – sensor data, quality control data, operational data, CAD / Content Addressable Memory (CAM) data, unstructured e-mail and social media data and so forth. The Gen AI journey will be accelerated in environments with ample, high-quality and well-integrated data. As the foundation to this, establish and support the connections in your organization between the business and process where Gen AI is being targeted, the technical teams steering implementation and the company’s data and analytics leadership.
4. Keep your eyes on the big picture
Gen AI is not the answer. It is part of the answer. Most large-scale Gen AI deployments are part of a wider program of transformation, including process optimization, wider automation, data analytics or combinations with other disruptive technologies. Manufacturing leaders need to understand how Gen AI fits into the broader transformation agenda and see it as part of this context, not as an isolated program that can succeed independently.
While there are dozens of use cases across manufacturing, their complexity and payback will vary significantly for different companies. Run a structured process to identify the initiatives in terms of cost / benefit for the first pilots. Selecting the wrong project could stall momentum and lead to failure.
Gen AI is a data-intensive technology, extracting new value from data in its many and varied forms – sensor data, quality control data, operational data, CAD / Content Addressable Memory (CAM) data, unstructured e-mail and social media data and so forth. The Gen AI journey will be accelerated in environments with ample, high-quality and well-integrated data. As the foundation to this, establish and support the connections in your organization between the business and process where Gen AI is being targeted, the technical teams steering implementation and the company’s data and analytics leadership.
3. Access the right capabilities
Gen AI success requires multi-disciplinary expertise, encompassing business analysis, domain expertise, AI modeling / programming / optimization, systems integration, data science and analytics, communications and project management. Where these skills exist in-house, nurture them. Where there is a major gap or need for speed, understand the partnership needs to assemble the right skills. In manufacturing, as for other areas of the economy, most organizations are learning that you can’t move quickly in AI by doing everything home-baked.
4. Keep your eyes on the big picture
Gen AI is not the answer. It is part of the answer. Most large-scale Gen AI deployments are part of a wider program of transformation, including process optimization, wider automation, data analytics or combinations with other disruptive technologies. Manufacturing leaders need to understand how Gen AI fits into the broader transformation agenda and see it as part of this context, not as an isolated program that can succeed independently.