AUTOMA Chem 2026: Key Takeaways on Industrial AI
Explore Cyzag's key takeaways from AUTOMA Chem 2026, covering industrial AI, trusted data, operational knowledge and smarter chemical manufacturing.
6 min read
A night-shift operator notices a temperature creeping up, adjusts the cooling and saves the batch. By morning, the reason behind that call exists only in their memory and a two-line handover note. Nobody did anything wrong. Most plants simply have no place to keep that kind of knowledge.
We were at AUTOMA Chem 2026 in Berlin on 26 to 27 October, and speaker after speaker came back to that gap.
Across presentations from Shell, Siemens, LANXESS, Evonik, Wacker, Endress+Hauser and others, one message stood out: the hard part is connecting operational information with the context and knowledge needed to make better decisions. Collecting more data or introducing AI comes after that.
This post is for operations and continuous improvement leaders in chemical plants who are being asked where AI fits. Below are the takeaways that stood out to us, a check you can run with your team tomorrow, and where Cyzag, an operational intelligence platform for asset-intensive industries, fits in.
The key takeaways from AUTOMA Chem 2026 centred on the importance of trusted industrial data, connected operational knowledge and human-led decision-making. Presentations from Shell, Siemens, LANXESS and other industry leaders highlighted how manufacturers can improve operational visibility, preserve expertise and establish stronger foundations for industrial AI.
Trusted data and context come before AI
Endress+Hauser emphasised that AI starts with trusted information from field instrumentation. While modern devices can provide detailed operational data, its value depends on connecting those measurements with the right context and delivering actionable insights to the people who need them.
During the opening panel, representatives from Linde, DSM-Firmenich and Evonik also discussed the importance of governance and human oversight. One challenge that stood out was the lack of structured records capturing operator actions and the reasoning behind operational decisions, leaving valuable knowledge disconnected from plant data.
This is the point we think matters most: the most valuable information in a plant is often the reasoning behind operator decisions, and it is rarely recorded. In a 2026 survey of 500 US manufacturing professionals, run by Dynata for Relay, only 20% said their formal systems very accurately reflect what happened on a typical shift. Four in ten said only some of what frontline workers discuss informally ends up in a system, log or formal report. Your historian shows what happened. It seldom shows why.
Bringing operational information to the point of work
Shell explored how operational information can be delivered to workers in the right context in its presentation, One Asset. Two Truths.
Its framework focused on four questions: Who am I? Where am I? What am I doing? And what do I need?
Rather than searching across multiple disconnected systems, workers should have access to accurate asset information, live operating conditions, approved procedures and relevant guidance when they need it.
Try this with your team tomorrow: pick one routine operator task and ask whether they can answer those four questions from one place, without opening a second system or calling a colleague. Where they can’t, you have found where knowledge is slipping.
LANXESS presented a similarly practical approach, demonstrating how QR codes can connect workers directly to the relevant asset records and applications.
Its existing digital environment includes separate tools for process monitoring, shift handovers, inspections and mobile work, highlighting the complexity of managing operational information across multiple systems.
This is what Cyzag does. It connects historian data, shift handovers, digital rounds and frontline workflows within one platform, accessible across desktop and mobile apps. Your historian stays where it is, and nothing migrates.
The opportunity is to make information easier for teams to find, understand and act on, rather than simply making more of it available.
Capturing knowledge before it disappears
Knowledge retention was another recurring theme, particularly as experienced workers retire and manufacturers face growing skills shortages.
Wacker outlined its path towards autonomous operations, placing people and processes ahead of data and automation.
Siemens highlighted a related challenge: while individual stages of the industrial lifecycle have been digitalised, knowledge is frequently lost between research, engineering and plant operations.
Its vision of connected models and workflows aims to reduce these information gaps, with one example demonstrating a 30% reduction in HDPE grade transition times.
In daily plant operations, the same loss happens from shift to shift. Valuable knowledge sits in shift conversations, operator experience and the reasoning behind decisions.
Capturing this information through structured handovers, digital workflows and continuous improvement processes helps preserve operational knowledge and makes it available to support future decisions.
From detecting deviations to taking informed action
Several presentations focused on moving beyond data collection towards practical, measurable improvements.
INEA’s discussion of predictive maintenance emphasised that intelligent monitoring is not simply another alarm system. By analysing multiple variables together, manufacturers can identify early deviations that might otherwise go unnoticed.
Importantly, these technologies were positioned as decision-support tools rather than replacements for engineering expertise.
Evonik also highlighted how operating constraints can leave significant value unrealised in chemical manufacturing, while Veela Systems emphasised the importance of measuring AI initiatives through tangible business outcomes rather than expectations alone.
That is the gap Cyzag is built for: finding the losses your alarms miss. Cyzag combines historian data, shift knowledge (logs, handovers and what crews know) and AI-driven root cause analysis. It is designed to show where a loss is building, what it costs and its likely cause, before it grows.
AI suggests; people decide. A person reviews and approves every action.
Turning industrial AI ambition into practical value
Throughout AUTOMA Chem 2026, a consistent message emerged: manufacturers need to prioritise practical use cases, trusted information and measurable outcomes rather than technology adoption for its own sake.
Shell advocated starting with one consequential operational decision, demonstrating value and scaling only when the evidence supports it. Siemens similarly recommended proving specific use cases against real plant data before expanding.
For Cyzag, this points to a practical way in. Cyzag Operations connects your plant data, shift handovers, rounds and workflows. Cyzag Intelligence is designed to add AI-driven root cause analysis on top of your historian data. You can start with either one, on one decision that matters, and expand when the results justify it.
Either way, the aim is the same: losses found earlier, their cost made visible, likely causes suggested and repeats prevented, with your team’s knowledge staying in the plant when people change shifts or retire.
The future of industrial intelligence depends less on what AI can do than on giving people the information, context and confidence to make better operational decisions.
Not sure where your plant should start? Take our Industrial AI Foundation Assessment to see how ready your data, workflows and knowledge are for AI.
Frequently asked questions
What were the key takeaways from AUTOMA Chem 2026?
Key themes included the importance of trusted operational data, connecting information across industrial systems, capturing frontline knowledge and introducing AI through practical, measurable use cases. Speakers also emphasised the importance of human oversight in industrial decision-making.
How can chemical manufacturers prepare for industrial AI?
Chemical manufacturers can begin by improving data quality, connecting existing operational systems and capturing the knowledge behind everyday decisions. Establishing structured workflows and reliable operational information helps create a foundation for more advanced AI applications. Our whitepaper on building the foundations for industrial AI covers this in more depth.
What is Cyzag?
Cyzag is an operational intelligence platform for asset-intensive industries. It combines historian data, shift knowledge and AI-driven root cause analysis to find the losses your alarms miss, with a person approving every action. Your historian stays in place, and you can start with Cyzag Operations or Cyzag Intelligence.
Follow Cyzag on LinkedIn for new articles and stories.


