We asked Australian miners to submit their biggest questions on digital twins and then gathered Willow’s resident mining and digital twin experts Tom Rivett and Rob Foster to shed some light on the answers.
Check out this blog to learn more on the basics of digital twins!
1. What’s the difference between a digital twin and software simulation?
In some contexts, the terms digital twin and simulation software may be used interchangeably, particularly if you’re looking at industries that have been using digital twins for a long time. Manufacturing and automotive are good examples of this, and digital twins are used to model parts and systems – before they’re even manufactured – to simulate performance.
In mining, digital twins are still finding their feet. In most off-the-shelf digital twin products available to mining, simulation is unavailable and where it is available, it’s being custom built for specific mine sites and use cases.
It is useful to consider software simulation as a tool for scheduling, optimization, process design, and CapEx planning, while a digital twin is used for operations and to highlight deviations from the simulated predicted capacities.
As mining twins mature, those design simulations may be handed down throughout the project lifecycle (from greenfield pre-feasibility to feasibility through detailed design and construction, commissioning, short term scheduling, ramp-up and back to upgrade pre-feasibility). This trend is already evident in spatial twins for major infrastructure.
Simulation is a capability we are building towards in WillowTwin for Mining.
2. Are businesses and leaders targeting too big of a picture when considering digital twins?
There is a lot of confusion about digital twins, and this is causing misalignment between the expectations of what a digital twin can achieve today vs what it will be able to achieve in the future.
There needs to be a balance in building towards the big picture and delivering value for today at a smaller scale. Trying to solve all the problems on site using one all-encompassing system is a risky endeavour and it’s advisable for everyone involved to start smaller. The final big picture won’t be clear until it’s tested and refined on site, over time.
Rushing this too early and quickly can have the opposite effect. IT does need to anticipate what the future state will be. This means they can build suitable architecture the first time. However, business users have specific problems that need to be solved which can be much smaller than the future state.
3. Where is the twin pulling data from, and why bother when so much of the physical world is not producing data we’re capturing?
There are two parts to this question but it’s a very important one.
The first part, where are digital twins pulling data from, is very specific to your site and systems. If you have IoT sensors installed on all your equipment, that’s a great start. A good digital twin will be able to integrate with any IoT hardware provider to ingest that data. The same goes for the systems you use to operate – SCADA, historians, EAMs, and all the other must-haves.
Where you’re not able to automatically ingest data, you can supplement this with a mixture of semi-autonomous and manual data collection. In some instances, this will look like a person manually inputting data into the twin, or it could be that person performing workorders or maintenance and the twin pulling that data over from other systems.
The second part of this question is ‘why bother’.
Surfacing relevant information to decision makers not only delivers value but is necessary – regardless of what form it takes.
Traditionally we have relied on mine site experts with 20+ years of tacit knowledge of the site to give accurate and valuable insights when making critical decisions. Although this is invaluable, it presents a bottleneck in the form of key-man-loss as these mining veterans are beginning to transition out of the workforce faster than they are being replaced.
A primary goal of a digital twin should be to enable this critical decision making to occur in lieu of tacit knowledge. As market expectations continue to grow, there must be progress made into exploring and ingesting real world data from the physical world. This is why WillowTwin is firm on being agnostic and flexible with respect to the physical data source. We are exploiting this approach with our Marketplace offering.
Furthermore, the OEM approach to twins is a more traditional product centric approach rather than a process centric approach. Their information is of value to reliability or design functions, however the scope for tangible business improvements around providing better data to reliability or design is limited by the maturity of those fields – they already do a pretty good job and can use transferable skills of experts from one mine to the next.
4. Is the twin for O&M teams or data teams?
WillowTwin is designed with O&M teams in mind. The goal is to make complex data more digestible and usable for those on the frontline.
Data teams are likely to require more granularity than is provided in the twin and they are equipped with the tools to dig directly into the existing data silos to get it. The twin will still provide context and prioritization. Nevertheless, there is still a place for data teams in O&M twins to make sure everyone is ‘singing from the same hymn sheet’.
5. Is a closed loop control system in a process plant a digital twin?
This is a flavour of a very common question we get which is, “is [this thing] a digital twin?”
There is no prescription as to how a digital twin presents itself. There are some criteria that can be mapped against something to see if it might be a digital twin, but it doesn’t have to meet 100% of those criteria if it’s not valuable in a certain use case.
For example, the digital twin of a crusher doesn’t need to be a 3D, animated model even though this image is what most often comes to mind. As long as the information is being presented in the format that makes it the most useful, it can be a digital twin.
Back to this particular question, yes, a PID loop meets the criteria of a digital twin.
Got a question that wasn’t answered here? Read up on the WillowTwin™ for Mining functionality.
Want to receive updates on new mining technologies and industry trends? Subscribe to our monthly mining newsletter!