Conundrum AI helps decarbonise cement production processes, including grinding, pre-heater and kiln, by utilising its powerful physics-informed AI engine. It enables considerable cost saving, yield boost and CO2 emission reduction. Our technology provides AI powered optimisation of processes that align with complex business objectives.
In an increasingly competitive market, embracing AI isn't just an option — it's necessary for growth and staying relevant in the industry.
Each solution consists of several parts: the process model, control optimisation policy, fuzzy logic as a fallback mechanism, and many other tool sets (data processing, analysis, etc.). Conundrum allows the use and combination of different types of AI models, specifically Deep Neural Networks (DNN), Recurrent Neural Networks (RNN) and Sparse Identification of Nonlinear Dynamics (SINDy), for the process model. This allows significant flexibility and achieves the best possible results from optimisation. For the control optimisation policy, we provide the same flexibility to configure various built-in methods, including Nonlinear Model Predictive Control (MPC) and Linear MPC.
Kiln Process Optimisation
The kiln process often has high material variability, environment changes factors, long process and indistinct and non-linear dependencies in the process.
This can lead to:
Using advanced process control strategies, Conundrum's AI solution stabilises temperature profile by optimising control of the kiln process, which enables lower energy consumption and consequently lower emission. Conundrum runs the most optimal processes to achieve optimal results:
Stylised representation of a rotary kiln
Unstable and high temperatures within the kiln
Coarse product with high variability in particle size distribution, density
Reduced product quality
High energy consumption and CO2/CO emissions
Between 10-30% Decreased Natural Gas Usage
+4% Improved Calcination
Up to 12% Decreased Energy Consumption
Stylised representation of a vertical roller mill
+2.5 average throughput increase
Between 5-8% decreased energy consumption
Cement Grinding Optimisation
The grinding process faces challenges such as high material variability, environmental changes, and difficulties maintaining the stable operation of the VRM, especially under varying load conditions.
This can lead to:
Inefficient energy usage
Variations in product fineness, affecting the cement's performance when used in concrete
Operational issues such as vibration, reduced grinding efficiency, and inconsistent product quality
Our process control solution maximises throughput in the grinding process while maintaining the desired level of product fineness and reducing energy consumption.
The platform's optimisation solution aims to achieve the following:
Throughput Maximisation:
The solution enables an average throughput increase of 2.5% in the grinding process.
Product Fineness Control:
The solution maintains the fineness of the ground, even as throughput is increased.
Advanced control algorithms and real-time adjustments uphold the desired fineness.
Energy Consumption Reduction:
Our control solution can reduce energy consumption between 5-8%.
By optimising grinding parameters and operating conditions, the system minimises excess.