Plants are designed once
Process knowledge gets locked into site-specific equipment, documents, and control logic.
THESIS_001 // CHEMICAL AUTONOMY
Chemical manufacturing must scale. The factory model cannot.
dominus is building toward autonomous chemical plants: systems that observe their own process, adapt within operating limits, and reproduce capacity closer to demand.
The chemical industry spent a century learning how to scale up.
Now it has to learn how to scale out.
01 / THE LANDSCAPE
One failure travels.
Feedstocks, conversion plants, shipping corridors, and end markets form one global machine. Chemical capacity is hard to substitute quickly, so a disruption in one region can surface several industries away.
NODE_07 / DEMAND + CONVERSION
East Asia
Large production clusters connect intermediate chemicals to electronics, batteries, pharmaceuticals, and consumer goods.
01FeedstockOil, gas, minerals, biomass
→
02ConversionPrimary and intermediate chemicals
→
03TransitPorts, canals, storage, shipping
→
04ProductionMedicine, food, energy, electronics
02 / THE PROBLEM
Conventional capacity is difficult to copy.
Most plants are singular projects. Their economics, equipment, control logic, and operating knowledge are tied to one site. Scaling usually means another long, custom project.
Operations stay local
Each facility rebuilds expertise through its own operators, vendors, and maintenance history.
Learning stays fragmented
Plants collect data, but they rarely turn every run into reusable operating intelligence.
The scaling penalty compounds
More capacity brings another custom design, another operating team, and another isolated learning curve.
03 / THE SOLUTION
Make the plant a learning system.
An autonomous plant is designed around a closed operating loop. It can read its physical state, make bounded decisions, execute them, and retain what it learns.
Observe
Turn physical state into trusted, continuous data.
Decide
Evaluate the process against models and operating limits.
Act
Coordinate controls, equipment, and operating workflows.
Learn
Carry the result into the next run and the next plant.
Sensing, software, equipment, and operations have to be built as one manufacturing architecture.
04 / THE SCALE
Scale the output and the footprint.
Chemical manufacturing has to produce more from existing assets and add capacity in more places. Autonomy provides a common operating layer for both.
MODEL_AScale up
- One larger site
- One custom learning curve
- Capacity stays concentrated
MODEL_BScale out
- Repeatable process cells
- Learning shared across the fleet
- Capacity deployed near demand
Increase useful output
Use a tighter operating loop to improve consistency, availability, and energy use.
Reproduce capacity
Deploy validated systems without rebuilding the entire operating model at every site.
05 / WHY NOW
The pieces arrived before the operating model did.
Closed-loop laboratories are already an emerging field. Plant sensing, robotics, process models, and industrial compute can now support a similar shift in production.
The lab is closing the loop
NIST describes autonomous laboratories that generate, characterize, and select samples with little human interaction.
NIST context ↗Every major industry needs chemistry
Energy, food, medicine, electronics, materials, and defense all begin with chemical conversion.
IEA context ↗Geography is an operating risk
Governments are mapping critical molecules, production sites, and external dependencies as strategic infrastructure.
EU context ↗The energy stakes are large
The chemical sector is the largest industrial energy consumer. Better operations matter far beyond one plant.
IEA context ↗06 / EARLY ACCESS
Chemical capacity should be easier to reproduce.
We are speaking with people who operate chemical assets, depend on critical molecules, or think deeply about industrial autonomy.
contactkennma@gmail.com ↗