The use of AI in the cement sector has moved firmly beyond the pilot stage but, as a new report from Augury shows, those with the greatest AI uptake are finding it increasingly hard to identify their next steps.
For the past four years, Augury, a leader in industrial AI, and Endeavor Business Intelligence have surveyed industrial companies to uncover trends in their use of AI as part of its State of Production Health series of surveys. The latest edition was conducted in March 2026 and released in July 2026. It received input from 501 manufacturing professionals in the US, Germany, France and the UK, with representations from Director, Vice President and C-suite levels.
The respondents’ companies, all of which had revenues of >US$50m and 84% of which operated at least five production sites, were from a wide variety of sectors, including food and beverages, chemicals, consumer goods, building materials, energy and utilities (new for 2026), wood products, metals and mining, pharmaceuticals, oil and gas and, last, but not least, the global cement sector.
Main findings
The 2026 survey found that AI usage had risen in all industrial sectors covered year-on-year. The main use case is predictive maintenance, with 57% of respondents saying that their company used AI for ‘Predictive Maintenance / Reducing Downtime.’ This was a remarkable 22 percentage points higher than 12 months prior, when only 35% used AI for this purpose. Manufacturers are moving toward predictive maintenance faster than any other AI application, specifically aiming to reduce downtime and upskill their workforces.
The second-most common use case was ‘Tracking Energy / Reducing Loss and Waste,’ with 50% of respondents saying their company used AI in this way. This had increased by 16 percentage points year-on-year from 34%. A close third, at 49%, was ‘Process Optimisation / Maximising Yield and Capacity,’ which had also been at 34% in 2025.
Cement is ahead in predictive maintenance
These three most popular AI use cases will be familiar to cement industry participants, which, along with other building materials sectors, is taking the lead in this area.
Indeed, the global cement sector was the most advanced user of AI for Predictive Maintenance / Reducing Downtime, with 78% of respondents confirming that their business used AI for this. Producers of other building materials were second with 74% adoption, with chemicals (72%) and metals and mining (69%) not far behind. Other sectors focused on other use cases. For example, energy and utilities has the highest prescriptive maturity level, while consumer packaged goods were more focused on supply chain management and logistics.
Measurable improvements
In addition to increased AI uptake, the 2026 survey shows that a higher number of respondents can now measure the benefits that AI implementation is bringing to their business. The changes are most pronounced for the three most popular use cases. 55% said that they could quantify the benefits of Process Optimisation / Maximising Yield & Capacity, up from 31% in 2025. 54% said that they could measure the benefits of Predictive Maintenance / Reducing Downtime, up from 30%. 45% said that they could track the benefits of using AI for Tracking Energy / Reducing Loss & Waste, up from 34% a year earlier.
Respondents said that reducing emissions was the biggest sustainability aim that they were able to quantify using AI, at 48%. Reducing energy consumption was second at 47%, with lowering carbon footprint third on 44%.
AI for upskilling
Looking to future use cases, respondents also identified AI, in combination with augmented, virtual or mixed reality hardware, as a key way to upskill workers in light of falling staff numbers and the loss of expertise through the retirement of experienced workers.
When asked if they agreed that advanced technology like the Internet of Things, generative and agentic AI, machine learning and extended reality would positively impact workforce upskilling efforts, 45% strongly agreed and 49% agreed. The total of these two responses is 94%, up from 87% in the 2025 survey.
Rising investment
A remarkable 31% of respondents said that they would invest ‘significantly more’ in 2026 than they did in 2025. 52% said they would invest ‘slightly more,’ meaning that a total of 83% would invest more than 12 months prior. Only 7% said they would invest less year-on-year.
A reality check since 2025
In contrast to the rapid rise in the uptake and the number of respondents reporting the measured benefits of AI use in industrial settings, the percentage who reported their companies as ‘very advanced’ at adopting AI to solve specific production problems dropped by 11 percentage points compared to 2025. This may be attributable to something of a ‘reality check’ that reflects increased realism among respondents about the limitations of their current AI programmes and the emergence of roadblocks to further development. It is interesting to note that C-suite respondents consistently rate their organisations’ AI advancement higher than those at Director and VP levels, a gap that may reflect the difference between strategic intent and operational reality.
When asked about the primary roadblocks to the adoption of AI tools, 47% of respondents cited poor data quality as the largest concern, a rise of 20 percentage points compared to the 2025 survey. 78% of cement industry respondents provided this response, the highest of any industry. Building materials was a close second, with 75%. This likely reflects that, the further along an industry or organisation progresses along its AI journey, the more clearly it sees where data gaps are limiting its future progress.
The second most concerning roadblock was process and/or operational hurdles (43%, up from 28%) and the lack of proof that a use case had been proven (39%). This was a new option for respondents in 2026, so there is no comparable data from previous surveys.
In the 2025 survey, a scarcely believable 33% stated that there were ‘no roadblocks to adopting AI.’ In the 2026 survey this figure had fallen to 1%, again demonstrating the growing realisation that AI implementation may not be as straightforward as it first appears.
Concluding remarks
Effective AI can help industrial companies become more productive and differentiate them from their competitors. The opportunity ahead is significant and considerable progress has already been made. With investment increasing, pilot phase success, and quantifiable impact, industrial AI is moving with more clarity of purpose than in any previous year, and that foundation is already being built on the backdrop of changing pressures.
Augury’s survey reveals that manufacturers are increasingly turning to the same tool to win: AI that predicts and provides the right knowledge to workers at the right time. The manufacturers who close this loop, between the machine and the person operating it, are the ones who will define what best practice looks like for their industries.
At the cutting edge, the supposedly conservative cement sector is showing other industries the way when it comes to AI adoption, particularly regarding maintenance, tracking energy and optimisation. The sector’s experiences as it navigates the challenges ahead will be instructive for not just the sector itself, but the wider industrial landscape.
Note
This is an edited summary of Augury’s 23-page report, available at: www.augury.com/collateral/the-state-of-production-health-2026/
About Elan Greenberg
Elan Greenberg is CEO of Augury, an Industrial AI leader that specialises in machine health and digital transformation for manufacturing. Appointed CEO in October 2025, he previously served as the company’s Chief Operating Officer (COO). Before joining Augury, Greenberg was COO at Flock Safety and General Manager at DoorDash. He is a US Marine Corps veteran and holds a Masters in Business Administration from New York University Stern and a Bachelors in Science from Cornell University.
