My Top 3 Focuses for Solid Operational Data.

I have been working in Mining Tech for 25 years. Since starting as a Junior Network Engineer (I still cant believe they hired me!) and working my way up to a Tech GM, I have learned a lot about what it takes to solve some of the biggest data challenges people face in mining.

In fact, they are a lot easier to solve than most people think. You just need to change how you're thinking about them. Here they are, and how to solve them:

1. Source System Data

Whether it's your Fleet Management System, Plant Control system, or Historian - we all know the pain of the source data not being correct! If your source data isn't right, you don't stand a chance of getting quality data for your decision-making process.

Typically, there are two focus areas here: comms and people.

• Comms: First, you need to make sure your network is solid. Poor comms has led to more issues than any other in my experience - invest now in comms and invest wisely in terms of expected future data requirements.Admittedly, this is much more difficult in the actual mine pit – get really obsessed with how to ensure coverage, and your source data will improve.
• People: There are a whole bunch of people processes required to get your data right. I typically find them ignored and/or not monitored or controlled in operations that suffer from poor data.Start to treat these processes as critical (they are!) and you will see positive change - report on them, just like the rest of your operational processes, they are very likely just as important.

Now you have great source data, job done, right? Not so fast, this is just the beginner level - mo' data, mo' problems!

2. Reporting Process

Wait...Surely with solid source data, reporting is easy.

Umm, kind of. You didn’t think all that reporting just comes together like magic did you? Well, I can assure you it doesn't...let's begin to dig 😉.

Like I said, reporting can be easy, but in most organisations I work with, it is far from easy and quite honestly appears to be purposely designed to be as painful as possible!

The point I am trying to make is, how do we get reporting so wrong, so often.

This unnecessary complexity I find is rarely intentionally done, and usually grows out of organic changes over years, always with the best intentions, of course.

However, this doesn't absolve us of our reporting sins, and we must repent. In the spirit of the last line above, look at the main causes of reporting hell!

Substance over Form - Incorrect Numbers:

"Harry, check out these amazing reports we wrote in Power BI; they are totally cool, right?"

"Yeah, they look amazing, there are just a few issues, do you mind if we have a chat about them?”

"First, they are wrong, see that field you used there, yeah that's the wrong one."

"Second, they are wrong - see how you calculated that metric, yeah the formula doesn't work like that."

"Finally, your code, brace yourself, it is so unreadable it would be easier to start from scratch than to fix all the other issues!"

Now, I would not be that harsh, or maybe I would, but when you think about the real issue here 👇.

People are running a real business off that data, making decisions, allocating resources, building strategies – it is critical to get it right.

So when the data is wrong, I do get a little emotional and antsy. It’s a hard truth for most to swallow.

The fix? Focus on making sure the source data is interpreted correctly, and your business rules are applied consistently.

And, if you want to write your own reports in Excel or Power BI, treat your reporting development like a software development process, because it actually is.

Multiple Reporting Systems - Different Numbers:

Manager 1: "Fleet availability was 92% which is a great result for the team."

Manager 2: "Sorry, I had your availability at 89%, which is okay, I guess."

Manager 3: "I used the raw source data and had my engineers clean it up a little, and we had it at 88.2%."

GM - Throws a chair against the wall and leaves the room.

Different systems have different interpretations of data, have different ways of calculating metrics, and sometimes even have different copies of source data!

If you ever have conversations about “which number” is right – then it’s already a mess, you know it and I know it.

Of all the operations where I have seen data actually work, they have what I call a “Single Source of the Truth.”

I am telling you this because I rarely see it; there is always a bunch of good excuses why, but they are mostly just that, excuses.

There should be no reason why there needs to be multiple views into your performance that show different values. None.

If you want to get the most from your data in terms of value - get yourself that Single Source of the Truth and get everyone on the same page.

Holy crap - You do all that...every day! – Painful Processes

Every so often, I'll meet with a client and get blown away by the complex processes they are using to get daily or weekly operational reports.

Which is pretty impressive, given the amount of data processes I have reviewed.

It usually includes uploads, downloads, copying this tab to that tab, clicking this button - but only when Person X has sent you an email saying it is all good. Ring Person Y and get them to refresh Power BI - oh, and be careful because this number here is always wrong; we were never able to work out how to change it, so we just ignore it on the reports.

Finally, Mine Superintendent A doesn’t like to do commentary, so can you find out why they are under target and make a few comments…

Does this sound fun?

Do you think the risk of error is Low or High?

I think you know, and I think you probably have a good idea of how the people executing it feel on Monday mornings. (Maybe you are one of them!)

The most painful thing is, it doesn't actually have to be that hard. I would bet that all the time and money being sucked into this black hole of despair could pay for an automated system ten times over.

What's amazing is the number of people who know this and just push on, greatly annoying their entire team - day by day - and never stopping to think how it could be fixed - once and for all.

Get a good system, make life easier for everyone, and have solid data you can rely on…strange concept, I know.

3. Actually using the reporting systems, you have.

Yes, really.

This might sound crazy, but I actually see it quite a lot; the client will have pretty decent systems, the issue is - no one uses them!

How can this happen, why would someone want to revert to Excel or Power BI when they have an awesome system that "does it all".

To be honest, it's mostly because it is tolerated - and every time it is tolerated, it undermines your Single Source of the Truth.

Would you allow someone to use their own Safety Management System?

How about a portable finance system they run off a USB stick to track their spend?

How about buy a new fleet of caterpillar trucks because they don’t like Komatsu?

– ok that last one was too far 😊-

You don’t, and it’s because those processes are deemed critical - Now ask yourself how critical it is to have solid accurate data for decision making in your operation...you're welcome.

So next time someone has some "ad-hoc analysis", or whips out an Excel spreadsheet, use that as a trigger to insist that the analysis be completed using verified data – pointing out the risk of acting on data with potential errors for bonus points - and brick by brick you will be building a culture that treats data as critical and with the potential to deliver endless value for your business.

See... Totally easy right...

What else did I miss? – comment below.