“I sincerely believe that if we want to attract women into mining and engineering, we need to start very young”

In the first of this special series of interviews focused on inspirational women in the mining technology space, I spoke to Dr Penny Stewart, managing director and founder of PETRA Data Science.

I first met Penny in 2017 when I read about the work that her team had done to implement machine-learning algorithms and reduce mill downtime at Newcrest’s Lihir mine in Papua New Guinea. I approached her for a telephone interview which led to a great piece for Mining Magazine, and we have worked together on many articles since.

PETRA has gone on to take the mining industry by storm, but despite the company’s success, Penny’s door has always remained open when I have a tech-based question. She is generous with both her expertise and introductions to contacts, and that’s why I named her as one of the women in mining that I look up to; Penny walks the talk when it comes to supporting other women in this sector.

Tell us about your background and training. What attracted you to the mining industry?

As a third-generation mining engineer born in Kalgoorlie, mining is significant part of my heritage and identity.

I remember going to work with my Dad on weekends in Mt Isa, and being impressed by the infrastructure, equipment, blast vibrations and shaft headframe that whisked people below surface. Dad was responsible for implementing the new million tonne mass blasting methods, and I remember finding the ground vibrations from blasting very exciting!

I was also very lucky that my Mum provided my sister and I with lots of Lego and other construction toys, and very early electronic maths games tablets to play with.

Later as a teenager, I was also very lucky to learn about engineering through go-karting (Dad taught me how to service the engine), and I read about how engines worked in our World Book encyclopaedia).

I mention all this because I sincerely believe that if we want to attract women into mining and engineering, we need to start very young. While mining technology is all about IoT-sensors, algorithms and software, the physical world of mining company’s assets are still primarily: the orebody, and the engineering machines (motors, gearboxes, drives, and control systems).

While there is no doubt that mining technology adds huge value, I believe that to be a world-class mining technology provider you need to be passionately interested in these physical assets, and what makes them productive, efficient and reliable.

What led you from mining engineering to a focus on data science, and ultimately establishing PETRA?

I spent the first six years of my career following a traditional mining engineering career path, including spending over a year working underground as an operator and shift boss.

I became very interested in how rock mechanics affects our ability to execute our mining plans safely. At the time, rock falls had contributed to multiple fatalities in Western Australian underground mines, and the industry was responding with new guidelines for ground support and an increasing number of geotechnical engineers on mine sites.

After a couple of years working as a production and planning engineer at KCGM, I decided to go back to uni and study rock mechanics and blasting as part of my PhD project on minimising dilution in narrow vein mines. At the time, a lot of engineers were super focused on physics-based modelling, and whilst I did a bit of that as part of my PhD, I found statistical analysis more powerful.

I was fortunate to learn statistics from the Julius Kruttschnitt Minerals Research Centre’s (JKMRC) world experts in applying statistics to mining data, including Professor Tim Napier-Munn, Professor Bill Whiten and Professor Geoff Lyman, all of whom generously supported postgraduate students with informal and formal learning opportunities.

Data science is the confluence of mathematics, domain knowledge and computer science to productionise mathematics. As an engineer with practical statistical understanding, I could see huge potential for data science in mining, and in June 2015 founded PETRA data science to extract value from mining data.

PETRA is not just about developing machine-learning and optimisation algorithms for mining, it is also founded on the fundamental belief that geological variability has a huge effect on our ability to execute our mining plans.

PETRA’s technical director, Dr Zeljka Pokrajcic, also completed her PhD at the JKMRC, and she was inspired by her PhD journey to cofound the not-for profit Coalition for Energy Efficient Comminution (CEEC). Together, the more we experimented and researched machine learning algorithms, the more convinced we became of the huge value in using machine learning to simulate and optimise processing by taking into account geological variability.

The word PETRA is derived from the Greek word for rock ‘Petros’, and rock variability is very much at the heart of all of PETRA’s algorithms – it’s something that those from manufacturing industries fail to understand has such a huge impact on mining value chain.

What do you love most about working in mining technology? Are there any aspects that you dislike?

I love seeing our algorithms deployed into production; preventing downtime, predicting and optimising processes all along the mine value chain. It really is incredibly rewarding.

Sometimes as an engineer, I am not the best business person, because I am more driven by results than profit. That said, in the long run, results create trust and, in a small industry like mining, where vapourware is running rampant, trust is increasingly important.

In answer to your question about what I dislike. I dislike the transactional nature of mining procurement. For example, it is common practice in mining for procurement processes to require multiple quotes, even for new technologies and business models.

In practice, this means that your new technology business model could be put out to tender to your competitors, who may never have considered providing the solution you are offering until invited by the mining company to participate in the procurement process.

These practices don’t make sense for the mining company either. When machine learning is deployed using existing platforms and systems, they typically have a payback period of less than a couple of months, and so delaying implementation in attempt to reduce the cost may actually destroy value.

How do you balance leading an internationally-recognised company with having a family?

I guess you could say that my husband and I are constantly juggling, and we don’t sweat the small stuff. We have three school aged boys, and so there is rarely a dull moment.

I travel much more than he does these days, but a few years ago he was away every second week. Fortunately, his new job doesn’t require much travel, and so it is much easier to manage than it used be.

Is it easy? No, it’s not. But, is it worth it? Absolutely!

Which other women in mining do you look up to and why?

I very much admire Michelle Ash’s courage for articulating a vision for mining that most people wouldn’t have the courage to talk about.

It’s so easy to be the cynic in mining, and knocking people who experiment with new ideas, and so I look up to people who have the courage to be vulnerable, and who risk established careers and livelihoods to work in the risky area of innovation and mining technology.

Alex Moss, the CEO of Canaria, and Tania Walter, the CEO of Obzervr, are female founders whom I greatly respect, and look to for inspiration.

I also very much admire Zeljka for her no-nonsense attitude that helps us cope with the ups and downs of running a mining technology business. Having Zeljka’s honest feedback has been very valuable, not just for me personally, but also for the business.

I see more women are reaching executive levels in mining leadership teams, but I haven’t really had the chance to meet, or work alongside these women.

I feel very strongly that the mining industry needs more women in STEM roles and particularly in positions of leadership. What are your thoughts on this, and how we can attract more talent from both inside and outside the sector?

I agree that it would be great to have more women in leadership in mining.

As we grow our team at PETRA, I have learnt that some women and other minority groups tend to understate their experience and abilities during interviews, and so we do a lot of referral checking and in-depth technical questions during the interview to manage this risk.

In addition to attracting the right people, I believe if you have a culture where people feel like they belong, irrespective of their age, cultural heritage, race, religion or gender, they will connect you to a wide range of other fantastic people.

When your own people are keen to introduce you to their networks, it’s a great sign that people feel like they belong.

For us, attracting people from outside of mining means attracting people with mathematics and data science expertise, and both these areas also have low representations of women.

In my opinion, there is no silver bullet, and so, like I mentioned at the start of this interview, let’s make sure young girls are exposed to the types of play and experiences that will foster an interest in the STEM skills required to succeed in the mining technology space.