29/06/2026
Is Your Feeds Your Feed Mill Ready for an AI Control System?
In the March/April edition, I outlined the successful application of AI in batching and pellet mill control. The natural question is: was it worth it — and why did it work?
It took two years and significant investment to develop, train, implement, and optimise the system. But the outcome was not incremental — it was transformational.
This was not about adding equipment. It was about unlocking capacity that already existed.
The Results
The numbers are clear:
Batching optimised to 90 TPH
Three pellet lines consistently delivering 18 TPH each.
No idle pellet lines
Reduced changeovers
Improved product quality and consistency
Meals hitting dispatch targets consistently.
Operator input for exception-based actions.
Capacity increased from 1000 to 1300 tonnes/day — with no major batch blending or pellet line plant additions.
The plant is now running close to its true physical limits — safely and predictably.
That is a 30% capacity gain without capital expansion.
If your plant is constrained, the question is not can you build more — it is:
Are you already sitting on unused capacity?
The Objective
The brief was straightforward:
Increase throughput, capacity, product quality, and meet delivery schedules.
This meant solving far more complex problems:
Keep pellet lines continuously fed.
Start and ramp automatically and autonomously.
Run consistently at maximum TPH.
Deliver seamless changeovers.
Optimise cooling without compromising quality, reduce drying and fines.
Produce meals without disrupting pellets.
Maximise batching TPH.
Remove unnecessary operator intervention.
The Real Constraint
Most plants assume their constraint is mechanical.
The reality, it is often decision-making.
Batching and pelleting are not static processes — they are continuous, multi-variable, real-time decisions:
What to make next
When to make it
How much to make.
How fast to run.
When to switch
Even exceptionally good operators struggle to consistently optimise all of this — especially across shifts, fatigue, and changing conditions.
So, plants drift into:
running below capacity “to stay safe”
overproducing to avoid risk
inconsistent performance shift-to-shift
For More details follow the next post :
