8 criteria for GenAI adoption in manufacturing: PwC

Manufacturers are excited about use-case applications of generative AI (GenAI)

Building and scaling any given GenAI solution can consume considerable time, capital and energy. This raises the stakes and, as with any new technology, can lead to losing propositions. GenAI, as with all AI solutions, carries risks and, therefore, needs to be applied responsibly.

Getting it right can bring many advantages, including cost savings in labor on repetitive or tedious tasks, product development, improved product quality and competitive strength, to name a few. One of GenAI’s advantages is that it can accelerate or enhance the digital transformation of processes, functions and business models by securely (as well as more quickly, easily and economically) accessing and leveraging a company’s unique data and intellectual property. Just as important, GenAI can be accessible to most employees, not just those with sophisticated technical skills.

Making the smart factory smarter

Most manufacturers have been deploying IoT-connected devices and machines to their shop floors at a rapid clip. Leveraging GenAI, though, will likely enable manufacturers to further build upon the increasing digital connectedness of their assets. We believe this could be carried out on numerous fronts by enhancing predictive maintenance capabilities, improving service and production logging, and amping up production efficiencies — even product design and quality control. We see potential GenAI applications on other fronts, including improving shop-floor safety, strengthening cybersecurity of the operational technology environment, and even making supply chains more resilient and agile.

Wading into GenAI wisely — and responsibly

Building and scaling any given GenAI solution can consume considerable time, capital and energy. This raises the stakes. As with any new technology, then, can lead to losing propositions. GenAI, as with all AI solutions, carries risks and, therefore, needs to be applied responsibly.

Getting it right can bring many advantages, including cost savings in labor on repetitive or tedious tasks, product development, improved product quality and competitive strength, to name a few. One of GenAI’s advantage is that it can accelerate or enhance the digital transformation of processes, functions and business models by securely (as well as more quickly, easily and economically) accessing and leveraging a company’s unique data and intellectual property. Just as important, GenAI can be accessible to most all employees, not just those with sophisticated technical skills.

What is GenAI and what is its potential?

Consumer GenAI tools such as ChatGPT and Bing Chat AI have swiftly entered the common vernacular. But what are they, exactly? GenAI is a branch of artificial intelligence that generates new content or data based on existing information. The foundation of generative AI is the large language model, an artificial neural network that is trained on vast quantities of data. With this knowledge base, generative AI can teach itself to emulate human writing and apply similar cognitive techniques to things like images, video, audio and computer code. GenAI can open opportunities to power workforce transformation at scale, democratize access to digital capabilities and skills, and free up time for workforces to concentrate on more strategic work.

Where — and how — to make GenAI work for manufacturers

Given that there may be potentially hundreds of prospective use cases, identifying the right ones for the greatest return is no easy task. And, as manufacturers venture to improve the experience of their factory-floor workforces, GenAI could play a role in enhancing roles and contributions of frontline workers, lending greater meaning and satisfaction to their work.

Manufacturers hold the opportunity to assess how to most effectively wade into GenAI use cases. We believe the first foray into GenAI ought to be a solid vetting process to help develop a rationale supporting what could be a considerable investment in time and resources.

We recommend the following eight criteria when assessing specific GenAI use cases.

Eight criteria for manufacturers to consider for GenAI

1. Determine return on investment

Does a given GenAI use case candidate solve an important problem/issue that will have a material and positive change for a company? Will it deliver a measurable ROI?

2. Do you know whether you possess the right data — and enough of it?

Organizations will need to ascertain whether their data is sufficient to support a GenAI use case, including whether the data is readily available, accurate and reliable. Also, will more useful data be incoming over time, or is it a static/fixed amount?

3. Hone in on automating lower-value tasks

GenAI can address work that is typically repetitive and time-consuming, so employees can spend more time on analysis and decision-making rather than on “manual” data collection.

4. Be realistic about what’s needed to scale up

To optimize GenAI, use cases should be scaled across business units to increase the ROI. Organizations will need to determine whether they have the right talent and capital resources to scale up. Just as important, they’ll need to prepare for the possibility that they’ll meet employee resistance that could impede successful implementation. Additionally, it’s critical that organizations consider the pros and cons to how it could affect their workforce.

5. Ascertain whether the data is ethical

Ethical data is data that is transparent, dependable, objective and unbiased. GenAI is only as effective and useful as the ethical nature of the data it draws from.

6. Make GenAI use cases intuitive, user-friendly and easily trained

According to the August 2023 PwC Pulse survey, 65% of leaders in the industrial sector say they are either already training employees on new technologies including AI and GenAI or have a plan in place to do so. As with any other new technology adoption, user-friendliness is critical and achieving that with GenAI is critical.

7. Integrate throughout the organization

In order to facilitate a successful scale up of a GenAI use case, it will need to be adopted throughout the organization. Doing that means creating GenAI solutions that are accessible and can be smoothly adopted by different business units which may have vastly varying digital skill sets.

8. Determine how a use case can evolve — and improve — over time

Ideally, GenAI use cases will become more powerful and efficient as more data becomes available to power them. It’s important, then, to assess whether it’s tenable to add more data continually in the future, and whether the data being added is high quality.