We covered few key topics that required a lot of thought. This group is tackling a really difficult topic in an industry that in the past has not really put in the effort.
There seems to be some issues on just getting off first base - i.e. agreeing on scope. There is an elephant to be consumed and it's difficult to know where to begin.
COMPELLING ISSUE:
It looks to me that, after joining several calls with this group, the most compelling issue that keeps coming up is the willingness of the OEMs to provide the raw data from their equipment that mining companies need. (OEMs = Original Equipment Manufactures - the companies who build and sell the mining equipment, such at Caterpillar, Komatsu, Sandvik, Atlas Copco, etc.)
The OEMs seem to think that the raw data being collected by their on-board monitoring systems, particularly that involving the health of the machine, is competitive IP and they do not want to part with it, even when the mining companies have purchased the equipment.
If the data is not even made available, then other issues such as data standards and transfer protocols are secondary. Having said that, I do understand that we could indeed just work with the production related data, which the OEMs are happy to part with, and still do something very useful. This includes things such as speed, weight carried, current position, etc.
However I think that we're seeing the rise of big data / advanced analytics solutions that are able to balance the reliability of equipment with the production task of the equipment. Miners do this already with preventative maintenance aimed at optimised production. They will need almost all the data to do it properly in the future.
Furthermore, in the case of new drilling equipment, we're also seeing data that was originally aimed at machine health monitoring being used as measures of the physical properties of the rock being drilled. Vibration data from trucks for measuring road conditions is another example.
Without naming names, I happen to know that a major OEM has been having this argument with their largest mining client. However, because this client only represents 3% of their business, then there is not a lot of leverage that just one miner can have in this particular argument.
This issue has to be tackled through broader industry collaboration. Perhaps there is a case that ONLY a group like GMSG can actually solve this problem. The sponsorship of the professional mining societies might prove to be key, as it has in the ore reserve reporting area.
SCOPE:
Where to begin eating the elephant? So in terms of scope, I think that ALL data being collected on mining equipment is in scope, regardless of the purpose, but certainly some data is more important that others. There is always some sort of 80/20 rule in place, whereby we should focus on the 20% that has the most benefit to mining companies.
The less important other 80% of the data (with only 20% of the benefit) will just have to fit in as best it can. Agreeing what is that magic 20% is the hard part, especially if we need to consider what's important for the future, i.e. data required for automated and remotely managed operations.
Another, point discussed yesterday was that unless we drill right down on an issue, we will not actually do anything useful. I agree, and I believe this is often best done by what I like to call the pi shaped project. It's a variant of the T shaped project (which is more often referenced in the context of the T shaped person).
In a pi shaped project, the scope covers a broad range at a high level (i.e. in our case: data access and usage across all mining equipment) and then we drill down on just two key areas to the level of depth required to actually get something done.
By doing two of these at a separated scoping distance, we avoid just solving a specific problem in a specific way that cannot be easily translated to other areas (hence the scope looks like the greek letter pi). That is, by doing two focus areas at once, we can see what issues are common to both as well as what is specific to each domain.
FOCUS AREAS:
In this topic of data access and usage, I would suggest the best focus areas may be:
- Open pit truck-shovel operations (not just trucks, but the shovels as well - i.e. two types of equipment that need to work together)
- Underground drilling and shot loading (and/or rock bolting) - to again get machines that need to work together.
Some of the more recent areas of application are in M2M (machine to machine) like collision avoidance, as well as H2M (the needs of humans to control the machines, like Human Machine Interface issues) and M2H (getting data back to the humans who need it). Sorry if I'm introducing unnecessary new terms, but that's just the way I think about the problem. (This working group itself is a H2H process for solving the problem!)
There also is a sort of multi-variate scope distance between these two listed focus areas that contrasts a few different dimensions. Open cut versus underground, earth moving versus drilling/blasting, low precision positioning vs high precision. There is also something special about how the Scandinavians have gone after this issue (in the underground space) versus the way it's been done elsewhere.
OTHER INDUSTRIES:
Another topic often raised is that we should look at how other industries have tackled this problem and hence learn how they ate their elephants.
I would NOT recommend we look closely at either the defence or aerospace industries for lessons on how to approach these standards issues.
When I worked for CSC (a global IT services company with major contracts in mining), I was often chasing down what we did in other industries to see how it could be applied in mining. Since the US Military forces together make up their biggest client, and NASA was CSC's oldest client, I ended up collecting many case histories from the defence and aerospace industries that were compelling and relevant for mining, including many that were managing lots of data from mobile equipment.
However these industries are each actually very difficult areas to get useable lessons from because they will spend whatever it takes over a long period of time to solve these data issues. Getting data off their mobile equipment (vehicles and satellites) is absolutely mission critical, so they've been tacking this for decades. So unless we want to spend lots of money over decades, then we can't follow the same path. They have now reached a level of maturity whereby the clients dictate the data and interface standards to their supplier, mainly through the use of open standards.
The Mars Rover program might be an exception - since it was done much more recently, very quickly from almost a standing start, involved many parties, and with a lower budget than usual. So worth having this one on our list.
However, the area that stands out to me as being worth a close look is airline industry, in particular the jet engine equipment suppliers to that industry. Like mining OEMs, they sell into a very competitive global market, with lots of different clients using much the same types of equipment. Their clients have an imperative to do things at low cost AND low risk. The way they have solved their data access problems seems to have been very effective.
For example, Pratt & Whitney and CSC worked together on a system for monitoring jet engines for a large variety of different civilian and military aircraft types. The advanced monitoring and analysis they have done has led to significant advances in both maintenance procedures and reliability based design.
As a result of developing such a monitoring system, Pratt & Whitney found they could manage the maintenance of the engines better than their airline clients, and so they now prefer to rent their jet engines as a service, including extending their monitoring system to handle data from the jet engines of other manufacturers.
Both GE and Rolls Royce also have similar jet engine monitoring systems and services, so it would be interesting to understand how that industry developed a much more open platform for data sharing.
As a result of these monitoring systems and the sharing of data, they have got into a virtuous cycle of using the data collected to continually improve both the design of the engines and the maintenance practices, and thus leading to more data being collected and analysed.
MEANWHILE BACK IN THE MINING INDUSTRY:
Instead we seem to be stuck in an industry where the OEMs make a lot of their money by selling parts into clients with a break-fix reliability mentality. Other than just keeping ahead of their competitors, what's the incentive for them to dramatically improve the reliability of their equipment? No wonder the equipment dealers are more interested in this problem, as they often have the maintenance contracts.
