The GenAI hype bubble has finally burst and it’s back to business as usual—worrying about paying out ransomware demands, finding the right people, decreasing supply chain reliability and figuring out how smart manufacturing helps with any of this.
Rockwell Automation’s 2025 “State of Smart Manufacturing Report” is appreciably shorter than last year’s. My suspicion, based on industry conferences thus far in 2025 and the almost desperate need by vendors industry-wide to make AI seem sexy in 2024, is that industry 4.0 has finally settled into “ordinary.”
The lack of perceived excitement around AI, part of manufacturing software stacks for quite a while, has nothing to do with decreased relevance. AI only became a starlet last year owing to the emergence of large language models (LLMs) such as ChatGPT (are we sick that word, yet?) that it turns out have little if anything to do with what takes place on the shop floor.
With the hype dissipated, manufacturers finally appear to be relegating AI to just another smart manufacturing tool in the toolbelt. No one gets credit for or gets to feel special for using AI, the challenge is figuring out the best ways to apply it.
I think industry 4.0 is at the same place and probably why we hear manufacturing experts trying to force discussions on industry 5.0 that sometimes feel downright ridiculous.
Smart manufacturing is the present, not the future. It means manufacturers have proven out the use cases and abandoned the failed pilots into purgatory. Now small and medium business can venture into the field without worrying so much about wasted CapEx and time.
Get Smart (Manufacturing) or Get Out
Most manufacturers (81%) say they are accelerating their digital transformation considering current challenges and “deploying and integrating new technology” took the top spot for internal obstacles, up from No. 2 last year. “Integrating smart manufacturing technology,” the No. 3 internal obstacle cited in 2025, was not even on last year’s top five obstacles list.
Last year’s report cited Cloud/SaaS as delivering the highest ROI. Respondents in this year’s report cite Cloud/SaaS investments as their top technology priority. The use case for abandoning on-prem servers and downsizing internal IT teams seems definitively proven, allowing technology teams to worry less about maintenance and patching and focus on innovation.
AI, predictably, took the No. 2 spot for tech investment priorities.
Speaking of AI…
Half of respondents plan in the next 12 months to use AI to support quality control. Last year, 45% cited quality control as their largest AI use case.
It feels as if manufacturers finally caught up with the early pioneers that deployed machine learning (ML)-based vision systems to detect anomalies in parts on conveyor belts passing under cameras, a decades-old application. We may have reached a point where manufacturers should no longer consider deploying this technology competitive advantage, rather than considering not deploying this technology a competitive disadvantage.
Barely less than half (49%) plan to use AI/ML for cybersecurity. Ironically, deploying AI tools alone exposes manufacturers to increased cybersecurity risk. According to the experts at IBM and Verizon, threat actors increasingly use AI to increase the effectiveness of phishing campaigns. Deployment of AI/ML becomes the new cyber arms race.
About two-in-five (41%) of respondents plan to introduce AI/ML to fill skills gaps and address labor shortages, upskilling existing talent and increasing productivity by shifting workers to more value-added tasks. One-third of respondents plan to use AI to help manage supply chain.
Smart Manufacturing Requires Smarter Humans in the Loop
Digital transformation equates to understanding old processes and transforming them into more efficient, digital equivalents. Most (83%) respondents in the 2025 report cite analytical thinking and communications/teamwork as the most important skills for new workers.
Just under half (47%) of respondents cite that applying AI represents an extremely important skill in their organization, up from 37% last year.
“Attracting employees with desired skillsets,” the No. 1 internal obstacle cited in 2024, sank to the No. 5 spot in 2025. This matches recent jobs data showing slowi wage growth and fewer open jobs in the manufacturing sector.
Last year, only 44% of respondents said they use data effectively. That number didn’t change in 2025 and neither year’s report puts a finger on why. Anecdotally, this does reflect recent conversations with manufacturing software providers about how clients keep asking for “data lakes” without knowing what to do with them in the first place.
You may download the full report here from Rockwell Automation.
