Cam Mounsey can see a future where every mining company on the planet is using machine-learning and remote sensing to monitor environmental impacts.
As the founder and chief executive of Tyton Ecological Intelligence, Mounsey leads a team of spatial ecologists, machine learning engineers and software specialists developing cutting-edge technologies
to drive better environmental outcomes across both new and existing mine sites.
Tyton’s signature product is a state-of-the-art machine learning model which can identify vegetation lifeforms, target genres and species at an individual plant-to-landscape scale. It is understood to be the
most advanced tool of its kind on the market and is powered by the company’s own ecological information system which conducts in-depth and precise assessment of vegetation classes.
The Tyton AI for monitoring and Tyton EIS for management end-to-end solution has been used by some of the biggest names in the industry, including Rio Tinto Ltd, BHP Ltd, Fortescue Ltd and Mineral Resources Ltd.
Mounsey said adoption of these types of machine-learning technologies in the field of ecology had rapidly increased over the past few years.
“I went to an environmental conference last year and there were probably half a dozen presentations on remote-sensing. Before that, it was rarely talked about,” he told Paydirt.
“We’re now seeing even some of the regulators asking about it, so there’s a bit of a trickle-down effect. Usually it’s just the bigger players who have the budgets and the appetites to support the R&D, but
I think the writing is on the wall because everyone knows this is the way it’s going.
“I think we’re not that far away from whole-of-site remote-sensing assessments being the standard method for environmental monitoring.” After completing his PhD in restoration ecology, Mounsey spent time in the industry working for one of the big iron ore miners. It was during this experience he identified what he described as a “real opportunity” for companies to improve the way they not only monitor the environment, but also how they manage data collected from the process.
In 2017, he established the environmental consultancy that would go on to become Tyton as it stands today. Within two years, the company completed its first whole-ofsite assessment using machine-learning.
While certain divisions of the Tyton business have been spun out over the years, such as Millcrest Environmental Technology and Spectrum Ecology, Mounsey said the company maintained the multidisciplinary approach which has been key to its success to date.
“The fact we’ve got machine-learning engineers sitting next to spatial ecologists, coupled
with our industry experience, is what’s allowed us to build this product suite that seems to be hitting the spot in the market,” he said.
“It comes back to the problem I saw when working in the industry and that was there’s lots of environmental datasets being collected, but rarely are they talking to each other. It’s worth remembering that ecology is the study of how organisms interact with the environment and these interactions are very important to understanding why things are going on.”
Mounsey described Tyton EIS as a “onestop-shop” for mining companies to manage all of their ecological data. “A task that might have taken days before is now essentially the click of a button,” he said. “Things like putting together an appropriate seed mix to rehabilitate the site could be quite a time-consuming task previously, just because of the way that data was sought or wasn’t collected in the first place.
With Tyton EIS, the system can interrogate your baseline flora and vegetation data at essentially the click of a button, it can recommend to you an appropriate seed mix and the quantities to sow them at and
hopefully return a vegetation community that resembles the local analogues while also meeting your completion criteria targets.
“Our model does a very good job at first-pass classification of saying ‘that’s a tree, that’s a shrub, that’s a herb’ so if you understand the structure of the vegetation community, you can infer a tremendous amount about the way that community is functioning and then track that through
time.”
At the time of print, Tyton was preparing to launch a service version of its Tyton AI product which can be readily used by regular ecologists.
“Machine-learning is a very technical subject and typically requires experts to carry it out, but the idea of the Tyton AI platform that we’re launching is looking to empower regular ecologists around the world to use this technology to conduct whole-of-site assessments, identifying individual plants,” Mounsey said.
“We think this is what’s going to revolutionise the way we monitor the environment, by putting these tools in the hands of the local experts.”
– Michael Washbourne, Paydirt
