Artificial Intelligence Image by Tung Nguyen from PixabayRootstock has published findings from a new report entitled “State of AI in Manufacturing”. The report is based on a survey of 350 respondents working in Manufacturing based in Canada, the UK and the US. Most of the respondents were from manufacturing operations rather than the front or back office functions.

The report highlights that 82% of manufacturers, despite other budget demands, are planning to increase budgets for AI over the next year and a half. Furthermore, 24% plan to increase budgets by 26%-50%.

The adoption of AI is emerging as a major priority for boardrooms worldwide. This survey confirms that manufacturers are also eager to adopt this new technology. It also looks at how manufacturers are using AI now and where they expect to apply it. The report is 22 pages long and has more PowerPoint slides than a whitepaper, with data visualisation on each page and a brief narrative.

Raj Badarinath, Chief Product Officer at Rootstock Software
Raj Badarinath, Chief Product Officer at Rootstock Software

Raj Badarinath, Chief Product and Marketing Officer at Rootstock Software, commented, “We are witnessing the dawn of a new era powered by AI, and manufacturers are eagerly embracing these tools as they’ve seen its potential to unlock powerful data insights across critical functions such as inventory, production planning, supplier collaboration, and more.

“While AI is not a new concept in the industry, we’re still just scratching the surface of what is possible. As our survey indicates, manufacturers need a solid ERP foundation in place to successfully leverage AI-powered tools.” 

The report is divided into four sections:

  • The current use of AI in Manufacturing
  • AI & ERP
  • AI’s Impact on Jobs
  • Future of AI in the Industry

Enterprise  Times also spoke to Badarinath about the survey.

The current use of AI in Manufacturing

This section looks at how manufacturers are using AI. While 70% have implemented AI, some of the responses reflect the demographic. The top three areas where AI is used are production, employee training, and customer service. Finance and sales do not figure at all. The most common use of AI is around automation, but generative AI is being used by 35%  of manufacturers already.

The report also looks at the geographic difference and the benefits manufacturers are seeing. As automation is the most used, it is no surprise that improved efficiency (44%) and productivity (42%) were the top responses.

The report is interesting on the perceived barriers to the adoption of AI. Internal skills (49%), integration (43%) and costs (37%) are the top three. Later in the report, data quality was cited as a major issue, with only 12% extremely confident about their data. I asked Badarinath about this.

“It’s a great pickup that you pointed out. I think data from a volume perspective is quite significant right now with enterprises. Data from a quality perspective, which is exactly what you need for training the (AI) models, still requires a lot of work.

“People are essentially saying, look, we got lots of data, we got a lot of data islands, we got a lot of data silos, it’s spread all over the place. But unless I clean it, I tag it, and I put it in a way that actually makes sense to train the models, it’s garbage in, garbage out, right? Which is what they were pointing to.”


Beyond the data quality issues, the survey looked at which technology works with AI. 47% of respondents believe that ERP is intrinsically linked to AI. Other technologies that will also prove important are Big data/Analytics (45%) and CRM (43%).

AI’s Impact on Jobs

Most respondents view AI as a positive for their role. Only 12% fear it will take their job, though 38% fear it will slow them down. Perhaps they believe the intelligence will interrupt them rather than assist them, rather like VAR is currently doing for football (soccer). Despite that, 76% said they’re somewhat, very, or extremely excited to use AI.

There is a skills gap, as pointed out by the survey. I asked Badarinath what skills manufacturers are missing. While there may be some AI skills required, such as prompt engineering, Badarinath pointed out that over the last few years, the technical domain knowledge required by manufacturing employees has risen. For example, to become a prompt engineer, they also need domain knowledge of additive manufacturing and CAD. So, how will AI help?

“When you introduce AI into the mix, it is going to do one of two things. One, it’s mostly going to be invisible because the goal of AI is not to create further skills gaps. I think there is a skills gap that exists in manufacturing in general that gets exacerbated. But not, I don’t think, AI is going to be reintroducing something new.

“Number two, if anything, what AI is going to help them do, our manufacturers or customers do is lower the barrier of understanding and decision making within a manufacturing enterprise. Why? The goal of AI is to drive outcomes, from the experience of the older demographic to the friendliness of the newer demographic, but generative AI is the bridge. To me, that’s the bit that’s missing in most manufacturing today, in my mind, from a skills perspective.”

Future of AI in the industry

Ai is here to stay and increase across manufacturing. 91% agree that AI is somewhat, very, or extremely important to the future of manufacturing. In Badarinath’s canned comment, he spoke about AI only scratching the surface of what is possible. I asked him what he meant.

He answered, “There’s a lot here that I believe, at least from a Rootstock perspective, that manufacturers can significantly expect. All the investments done to date, especially around generative AI, have been on the front office side. Anything to do with unstructured text has been the main focus. I believe there’s still a significant part of structured data that still needs to be combined with generative AI to create some very interesting use cases.

“For example, take a look at the changing demographic of workers that we currently have in manufacturing. There’s a bunch of folks who are retiring, and there’s new guys who are coming in, which is a completely different generation, a different demographic altogether? How do you help train the new generation with the experience and judgement of the ones that have been running this business for 20, 30, 40 years?

“This is where a combination of generative AI, which is all about making the conversational interfaces a lot easier with the predictive parts, which tells you what the decisions are and what are the recommendations that need to be made, especially as you discover exceptions, I believe that’s where the true magic is. Frankly, none of the technologies today have quite cracked the code on the manufacturer’s side.”

Why does this matter?

Stu Johnson, Rootstock’s Vice President of Product Marketing, concluded, “There has been enormous pressure on manufacturers to find new ways to improve productivity and increase efficiency in their operations, so it’s satisfying to see how AI is accelerating those initiatives. As the available capabilities and integrations involving AI continue to advance, we’ll see a clear division between the leaders and laggards—with early adopters dominating their respective manufacturing segments.”

I asked Badanarinath what the findings of the survey mean to Rootstock. He replied, “There are three big things that I took away from this. Number one, you’ll see money going into AI. Clearly, there are a lot of expectations of growth and efficiencies, which is why people are prioritising investments in this.

“Number two, ERP is a core solution for AI. Remember, a bulk of the data that drives the decisioning sits with ERP. Today, we’re talking about generative AI, which is all about unstructured text. How do I write an email to somebody? For manufacturing, you really need the core data that helps them run their business better. ERP tends to be core with that, so that helps  Rootstock.

“Number three, think about the barriers that came out in the survey as well. One of the barriers was the lack of internal expertise, and another one was integration. The third one was implementation cost. When you pick Rootstock, which is on the Salesforce platform, the idea of leveraging the Salesforce skills into Rootstock and using that actually helps them to lower the barrier there from the expertise perspective.

“There is no integration. We are built into the Salesforce platform. As a consequence, there are no crazy implementation costs. It’s all about configuration. For us, we lower the friction of adoption of AI; that’s what makes it so exciting for us. We have the data, and you need ERP to make better decisions. Just being on the Salesforce platform gives us a differentiation that we need.”

Enterprise Times: What does this mean

For a fairly light report, this was well put together and offered insights beyond what many other surveys have done.  There was no breakdown by country or subsector, and the balance of respondents did skew the results, though it made them no less meaningful.

Talking with Badarinath added more depth to the results. The report did miss a qualitative element that might have added additional value. There were no quotes within the report, simply descriptive text and a light analysis of the quantitative results. However, for manufacturers, this makes an interesting read, and certainly, if the report is repeated next year, it may highlight some interesting trend data.


Please enter your comment!
Please enter your name here