Molycop

Operating Systems

Fundamentally, the objective of operations management will involve the management of the production of a maximum volume of units in the shortest time, at the agreed quality, for the lowest cost. The deliverables must include safety maximization and environmental impact minimalization. It is not achievable ad hoc. A rigorous, mechanistic, and explicit operating system is required.

Two such systems (there are others) that have been widely adopted, deployed, and learned throughout the operating world are LEAN and Six Sigma. LEAN began with learning from the Japanese Toyota car production system. The seminal published work is a book by Womack, Jones and Roos called “The Machine that Changed the World .”It was published in 1990, thirty-four years ago. LEAN learning has not stood still. LEAN declares five principles: defining value, mapping the value stream, creating flow, using a pull system, and pursuing perfection. One learns that LEAN thinking is a ‘war on waste’ within the operating system. Waste categories are explicitly defined. As are all work operations, procedures, processes, process times, and process improvements. Measure, measure, measure. 

An operating system based on Six Sigma teaches one the power of measurement and analysis. Focus on reducing variability and improving process capability. The seminal published work is a book by Forrest W. Breyfolge III called “Implementing Six Sigma.” 

The theory of constraints (TOC) is primarily used as an improvement strategy to maximize benefits in a value chain. The seminal published work is a book by Eliyahu Goldratt called “The Goal .”It was published in 1984, forty years ago. TOC has five principles: identify the goal, identify the constraint, exploit the constraint, subordinate operations to the constraint, de-bottleneck the system, and find the new constraint. Iteration after iteration of this process generates a cycle of continuous improvement and increasing efficiencies. Each iteration is mapped by a Current Realty Tree, a Conflict Resolution Diagram, and a Future Reality Tree. Measure, measure, measure.

It might seem that these operating strategies are ‘competing’ philosophies. They aren’t. 

Both have a tremendous amount in common. 

Both focus on a culture of continuous improvement, root cause analysis, elimination of inefficiencies, and a lifetime of assiduous commitment to the cause (not a fad). For a good comparison, read “Theory of Constraints and Lean Manufacturing: Friends or Foes?” by Moore and Scheinkopf, 1998. 

Process Control

At the same time that operating philosophies were developing, circa 1985, researchers and practitioners in the process control world were exploring the interaction of fuzzy set theory and expert systems in a construct called artificial intelligence. At that time, AI had the meaning of mathematical modeling and computer simulation of decisions that replicated people who were experts in their roles. A good summary of that journey is provided by Gaines and Shaw from the Department of Computer Science at York University. The article provides the principles of Expert System process control:

1. thoroughly instrument the system to be controlled or about which decisions are to be made

2. use the instrumentation to gather data about the system's behavior under a wide variety of circumstances

3. from these data, build a model of the system that accounts for this behaviour

4. from this model, derive algorithms for decision or control that are optimal in terms of prescribed performance parameters

In the thirty-to-forty-year period since, technological improvements in sensors, computing, and human-machine interfaces have resulted in step changes in Expert Control System capability. Combining operating philosophies and technology in operations has seen profound changes in the manufacturing industry worldwide. 

Generically, manufacturing systems are explicit, mechanistic, and controlled in real-time. However, some industries find it difficult to maintain adherence to the fundamental principles of an expert control system. 

This is especially the case in situations where plants are large, complex, or uncertain, or there may appear severe changes in operating conditions. We have just described the mineral processing circuit.

Monov, Sokolov, and Stoenchev talk about this in their paper “Grinding in Ball Mills: Modelling and Process Control,” 2012 (a decade ago). 

They say, “The process control in a ball mill grinding circuit faces severe difficulties due to the following well-known characteristics: the process is nonlinear with immeasurable disturbances and unmodelled dynamics; there are strong interconnections among variables so that each input variable interacts with multiple output variables; the time constants of the process have values in a wide range, and there are significant time delays in some input-output pairs; the system model contains a number of integrators; the process parameters vary in time as the circuit ages; there are technological constraints on the manipulated and controlled variables; the measurements are unreliable and noisy.” Variability is the insurmountable villain.

The Future

Using Moore’s Law of computing power as a rule of thumb, it is inevitable that AI will close the variability challenge gap in process control of tumbling mills. Supporting the computing power will be technologically advanced sensors that are robust, reliable, and repeatable in their data streams. At the same time, philosophies like LEAN and TOC will result in a demand for explicitisation and mechanistic processes, including a war on waste, that erode the variability of the process. Measure, measure, measure. Consistently, assiduously, and without variability.

The mineral processing teams need only to look at best-practice manufacturing facilities to see where the future lies.