Interview: Cemalettin Kanaş

Volatile raw materials, rising energy costs and tighter quality demands are putting feed mills under pressure. Andritz outlines how predictive control can help operators stabilize drying, reduce energy use and improve yield.

Robert Sabljo
Advanced Process
Control Engineer
ANDRITZ FEED
AND BIOFUEL

Drying remains one of the most energy-intensive and variable stages in feed production. Conventional PID-based control systems often struggle with long process deadtimes and frequent changes in operating conditions. In this interview, Robert S. Sabljo, Advanced Process Control Engineer at ANDRITZ, explains why drying has become a critical bottleneck for many mills and how ACE (Advanced Control Expert), powered by BrainWave MPC, is changing approaches to moisture control.

THE CHANGING LANDSCAPE OF FEED MILLING

Feed milling is under pressure to become cleaner, more efficient and less variable. From your vantage point, what fundamental shifts are you seeing in modern mills today?

Feed milling today is facing a perfect storm. Volatile raw materials, rising energy costs, tighter quality demands, and ever-thinning margins collide in the same production lines. As a result, mills are accelerating their move toward automation and digitalization, with a strong emphasis on data-driven optimization. Predictive control technologies are increasingly used to achieve optimized and efficient production while keeping product quality equal to or better than before, reducing manual intervention and improving overall process stability. At the same time, sustainability and adaptability have become key priorities, with mills aiming to lower energy usage and environmental impact while remaining flexible enough to handle frequent changes in recipes and raw materials.

Where do plants typically lose the most efficiency in drying and moisture control, and why has this stage become such a pronounced bottleneck across the industry?

Drying is one of the largest consumers of energy in a feed mill, often accounting for up to 60 percent of total usage. Yet despite its importance, variability remains the industry’s biggest headache. Overdrying wastes energy and reduces product weight, while under-drying risks microbial growth and damaging market reputation. The root cause is rarely the dryer itself. Instead, inefficiency emerges from conventional control systems that react too slowly, relying on operators to make manual corrections after deviations occur.

What specific operational pain points are pushing mills to consider advanced control solutions like yours?

Moisture variation is not just a quality issue but a cost problem. Product inconsistency driven by moisture variability leads to wasted energy, reduced yield and lost production capacity. To protect quality, mills often dry conservatively, sacrificing product weight and increasing energy consumption. Advanced control systems address these pain points by stabilizing moisture, reducing energy usage and minimizing operator intervention, delivering direct financial benefits

FROM REACTIVE TO PREDICTIVE DRYING CONTROL

For readers who haven’t seen your ACE (Advanced Control Expert) in action, how would you describe the system’s core function and what makes it fundamentally different from a conventional dryer control setup?

At its core, ACE functions as a “virtual expert operator”, continuously optimizing drying, predicting disturbances, and reducing manual intervention. Unlike conventional PID loops, which respond to deviations after they are detected, ACE uses BrainWave Model Predictive Control and a digital twin of the drying process to simulate future behavior. Inline moisture sensors feed real-time data into the model, allowing corrective action to be taken before errors show up.

Model Predictive Control isn’t new in heavy industrial processes. Why is MPC becoming relevant for feed milling now?

Feed milling demands flexibility. Recipes change frequently, production rates fluctuate, and PID controllers often struggle with long deadtimes such as dryer moisture control. In a drying process, a temperature change may only affect moisture readings 25 to 40 minutes later. MPC overcomes this limitation by anticipating process responses and adjusting conditions before deviations occur, making it particularly well suited to multi-zone dryers and slow-response systems.

Many suppliers now talk about “AI” or smart control. What differentiates BrainWave MPC?

BrainWave MPC relies on robust mathematical process models that are continuously corrected using real sensor feedback. It combines predictive modeling with advanced feedforward control to suppress measurable disturbances, such as upstream water addition, before they impact the dryer. This proven approach allows the system to remain stable even when conditions vary, without over-correction or excessive actuator movement.

REDUCING VARIABILITY AND HANDLING REAL-WORLD DISTURBANCES

You claim that the system reduces moisture variability significantly. What makes this possible, and what improvements are realistic?

ACE guarantees a 30–70 percent reduction in moisture variability, depending on recipe complexity and the quality of existing controls. By stabilizing moisture, mills can operate closer to their true target, increasing product weight and improving yield. In many cases, this allows the moisture setpoint to be raised safely while still meeting quality limits. Energy usage is also reduced through avoidance of overdrying and fewer actuator movements compared to PID control.

Mills rarely operate under stable conditions. How does the system adapt to recipe and material changes?

BrainWave adapts to that daily instability by using multiple scenario-based models that switch automatically when recipes or operating conditions change. Measurable disturbances are managed proactively, keeping the dryer stable and allowing operators to focus on oversight rather than constant manual adjustments.

How does ACE handle sudden, unexpected disturbances?

Sudden jumps in moisture are a daily reality in feed production. BrainWave models the variables affecting the system as feedforward inputs, allowing corrective action to be taken before disturbances reach the dryer. When disturbances cannot be measured, the system reacts immediately, minimizing their impact. Rather than chasing errors, predictive control prevents them.

INSTALLATION, INTEGRATION AND OPERATIONAL SECURITY

What should plants expect when adopting ACE in terms of installation and downtime?

ACE is designed for seamless installation. Moisture sensors are installed and calibrated first, followed by sampling and commissioning. The system is typically commissioned over roughly two weeks, with only minimal production interruption. Operators receive training, and remote optimization ensures the system remains fine-tuned after startup.

How easily does ACE integrate with existing or older plant control systems?

ACE overlays existing PLCs without replacing them. Operators can deactivate the system at any time, instantly reverting to legacy control. OPC connectivity enables smooth integration, even in plants with mixed equipment brands, allowing mills to adopt predictive control without a disruptive rebuild.

What fail-safes are in place if something goes wrong?

ACE is built with redundancy in mind. If any anomaly occurs, control automatically falls back to the existing plant control systems while alarms notify operators instantly. Production continues uninterrupted, ensuring stability is always maintained.