Smart sensing for smarter mines

Dr Simit Raval leads the Laboratory for Imaging of the Mining Environment (LIME), which sits within the internationally renowned Australian Centre for Sustainable Mining Practices. In this Q&A, he explains how smart sensors and drones are taking mine environmental management and safety to a new level, and why the time is right for the mining industry to embrace the smart revolution.

You’ve coined the phrase “ecosystem fingerprinting” to define one area of your work. Can you explain what it means?

Ecosystem fingerprinting is a means to characterise a native environment to aid its conservation and restoration. This involves using smart techniques and sensors to assess and monitor sensitive ecosystems, so looking at vegetation for its species composition, structure, bio-physio-chemical content, etc. We get the information by putting sensors on a UAV or drone and flying over the nominated site. It’s a new way to monitor and protect the environment being mined – for example, assessing the impacts of mining on the environment or continuously checking a mine site once it has closed.

Dr Simit Raval

Although my main focus is mining, ecosystem fingerprinting is also applicable to other industries such as agriculture and forestry. I was recently approached by an agronomy company to fly our sensors over vineyards and other crops in parts of Queensland and New South Wales. They found it so useful they now want to collaborate further: to use our system and algorithms to assist them with precise agriculture.

How did you get into this research area?

It started with my PhD in 2008, when I was using satellite-based observations to look at the impact of mining on the environment and assess mine rehabilitation success. We got some interesting results but the data was too low-resolution to be all that useful, so we started putting sensors on aircraft instead. This was successful but really expensive so we started looking at drones and UAVs. They’re now the main focus of our research. Ultimately, we’re looking to create an integrated monitoring system that involves satellite-based observations at a regional scale and drone-based monitoring at a local scale, and contribute towards sustainable mining practices.  

Can you give an example?

One study we’re running at the moment is monitoring a wetland/swamp area that lies above an underground mining operation. It’s quite a sensitive, heterogeneous and very complex ecosystem that’s home to a few endangered flora and fauna species.

We flew in with cutting-edge tuneable hyperspectral sensors that are able to provide information on the photosynthetic activity of vegetation to map its subtle health condition. The images we get are incredibly high resolution (one pixel of the image is 2x2cm of the territory) and we can pick up a whole lot of different characteristics of the target from what we call the “spectral signature” from that hyperspectral sensor. 

My vision for a mine site is to have a number of drones making automated observations that feed into a smart system to predict areas of concern and mitigate against them.

Dr Simit Raval, Leader of the Laboratory for Imaging of the Mining Environment

Can you also take water samples?

Yes. Last year, one of my undergraduate thesis students developed an ingenious water collection device that we trialled successfully in a mine tailings area (an unsafe environment from which it was impossible to collect samples). It’s a kind of remotely triggered vacuum pump we can lower into the water, and then suck the contents into a container on the drone for analysis.

The success of this technique has spread quite quickly and Sydney Water has recently asked us to explore possible collaboration on a project to do guided sampling remotely in real time. This means we’ll immerse some sensors into the water to measure quality, and then use that information to tell the drone where to take samples from.

Environmental monitoring is just one of the things your lab is working on. What else are you looking at? 

Mine safety is an equally big one. I’m currently leading an Australian Coal Industry’s Research Program project in which we’re putting LiDAR on a UAV, and using it to observe pit-wall stability in an open-cut mine. The sensor allows us to look at the structural parameters of walls in very fine detail. The next steps in this area of research will be to introduce machine learning and artificial intelligence to analyse the data and automate that process so that a geotechnical engineer can keep track of changing structural stress and better estimate possible failures.

Another sensor I’m using is an underground laser scanner that is based on an algorithm called SLAM. This gives us the accurate position of every location in a point cloud, even if we don’t have GPS coverage, such as in an underground mine. I’m using this system for mapping, positioning and monitoring underground deformations.

What direction would you like to take LIME in the future? Do you have any big plans?

Yes, I have big plans! At the moment, LIME is the only laboratory in the world that’s focused on using remote sensing and smart tools exclusively for mining, so we’re working in a unique area. I want to maintain that focus and become the global leader in that space. My vision for a mine site is to have a number of drones making automated observations that feed into a smart system to predict areas of concern and mitigate against them.

I think the time is right for these new methods and there’s a very good case for the mining industry to adopt them because they will be more accurate and more cost-effective.

Why is the time right?

Technological advances over the past six or seven years have made so many more things possible, and the systems continue to advance at a rapid rate. Not only do we have access to increasingly powerful, lightweight and cost-effective sensors but, simultaneously, we have access to increasingly powerful drones and UAVs with a higher load-carrying capacity. On the other hand, algorithms to handle the generated data are getting smarter, too. What is necessary is to intelligently mould them for specific applications.   

For more information

If you’re interested in learning more about LIME’s work or discussing a project, please contact Dr Simit Raval: simit@unsw.edu.au  

Share this