Many people that I interact with tell me that they want to use advanced analytics, big data, or even cognitive computing in their business. I’ll open by saying that these are all very powerful tools and will change the way that industries and society more broadly works at an ever increasing rate. But the emphasis is on tools. Analytical techniques, advanced or otherwise, are not solutions, and for the most part they also don’t identify problems.
The application of analytics should be done in a manner which is consistent with the problem being solved – shooting mosquitoes with missiles might work (not talking from personal experience) but would be a costly and over-engineered approach that would require a high level of expertise to set up, execute and maintain.
In a more conventional scenario, saying that ‘I need advanced analytics at my mine’ is akin to saying ‘I need another truck at my mine’ – depending on my frame of reference, I might assume that you were talking about a dump truck, a cement mixer or a maintenance utility vehicle. The questions that should follow are “what do you need to do with it?”, “why do you need to do that?” and ‘what are you trying to achieve?’, closely followed by “what other options have you considered?” and “how will this impact other parts of the value chain?”. There are a lot of options on a minesite to improve productivity that don’t involve buying another truck, and in some cases more trucks will result in lower overall productivity.
Purchasing another truck with a random configuration and inserting it into the minesite and then hoping that it works would have much the same impact as purchasing “advanced analytics” services or platforms without first defining ‘why’ – you might get it right, or you might end up with something that kind-of works but wasn’t worth the investment, however from my observations, you’ll probably end up spending a lot of time and money, get very frustrated with the analytics provider and either end up with nothing that works, or a trivial application in a super-complicated system.
Many companies seek to sell standalone advanced analytics services across a range of fields, however to consistently improve productivity using analytics it is necessary to follow a structured process:
- Define what you want to achieve in your business
- Describe what you need to do to get there
- Describe what’s stopping you from doing what you need to do
- Define the technical problems
- Shape the technical problems into mathematical problems
- Use the simplest approach possible to completely solve the mathematical problems (note that I’m not advocating creating a spreadsheet so complicated that only one person can use it – its about selecting the tool that is fit for purpose!)
- Maintain the deployed solution.
Following this process effectively requires a collaborative multi-disciplinary approach that will require cultural change on most mine sites, but the reward is a plethora of incremental productivity improvements. Most of the opportunities identified can and should be solved conventionally – large minesites often achieve tens of millions of dollars in annualised cashflow improvements before describing the first problem that actually requires ‘advanced analytics’. That said, sites that follow the process to ultimately achieve a level of maturity that can describe a complex problem will be well positioned to shape it effectively into a mathematical problem, and then to solve that problem. Over time, these companies will reap the rewards both from the step change in their business relative to their peers, and the highly valuable incremental work that they executed on the way through.
In my experience, you may passionately believe (quite rightly) that advanced analytics can change your business for the better, you may even have data, platforms and information technology architecture that technically enable advanced analytics, but if you haven’t defined problems that need advanced techniques, then you’re shooting mosquitoes with missiles.
Director of Operations, Interlate