Created by Sam Johnson, Geotech Lead
The Australasian Tunnelling Conference in Perth was a good reminder that the hardest part of tunnelling is still the ground itself. I went to share how we demonstrated that Datarock Core can turn core photos into consistent geotechnical data using the Sydney Metro West project as an example, and to learn how our mining tools can support civil tunnelling projects. After three days of talks and conversations, it’s clear there’s a real appetite for better ground data and practical tech that helps engineers make clearer, more confident decisions.
The Australasian Tunnelling Conference 2025 in Perth reminded us how far tunnelling has come, and how much harder it’s getting. The “easy” projects are long gone. Today, tunnels weave through dense cities, deep rock, and unpredictable ground, where a conservative engineering decision can cost millions, and wrong assumptions can cost much, much more.
For Datarock, this was our first real step into the civil tunnelling world. Until now, we’ve only worked in mining and minerals exploration. However, we wanted to see whether the same tools that help engineers and geologists understand ore bodies also help civil engineers make better decisions underground.
After three days of presentations and conversations with contractors, designers, and asset owners, we’re confident the answer is yes.
The challenge everyone feels: the ground still surprises us
Professor Jenny Yan opened the conference with a fantastic keynote about the engineering associated with the 66,000 kilometres of traffic tunnels that have been constructed in China over the last 30 years. Unsurprisingly, she noted that the biggest challenge is still geological uncertainty. Not money. Not technology. The ground itself. It’s not surprising because it’s the one aspect of tunnelling we can’t control.
Whether in mining or tunnelling, the reality is the same: we drill relatively few holes to get intermittent snapshots of the sub-surface, and then make big calls from this limited data. While that’s a reality of the industry, we can and should get smarter about how we use what we have. More reliable and consistent data means better design decisions and fewer surprises.
How Datarock fits in
In the Digital Engineering stream, I shared how Datarock Core could be used on the Sydney Metro West tunnels to quickly turn thousands of core photos into consistent, measurable, and auditable geotechnical data. When making decisions based on interpreted ground conditions, objectivity matters. For the presentation, I focused on Rock Quality Designation (RQD) because it is a very commonly logged dataset, and core dominant colour because it can often reflect key geological features that are not always logged by civil and geotechnical engineers. That said, these are just two of many datasets we produce as part of our Structure and Geotech solution.
Working with large mining datasets enabled us to reliably extract insights from imagery. Bringing that same thinking into tunnelling helps bridge the gap between what we know and what we wish we knew about the ground ahead.
A reminder about people
At the conference dinner, Arnold Dix gave an unforgettable talk about the Silkyara Bend-Barkot tunnel rescue in India. If you haven’t heard the story, I highly recommend watching that episode. During his talk at the dinner, his message was simple but powerful: no matter how advanced the technology gets, it’s still people who make the difference. Courage, empathy, and judgment matter most.
That really resonated with us. We build tools that make use of recent developments in AI, but what earns trust is the people behind them: the engineers, the support teams, the customers who use them.
Looking ahead
Tunnelling projects are only getting more challenging and the cost of poor, or overly conservative, engineering assumptions keeps climbing. Mining and civil engineering still have a lot to learn from each other. Tools like Datarock Core aren’t silver bullets, but they do help engineers trust their data more, which means designs can be smarter, more efficient, and ultimately more cost-effective.
The Digital Engineering stream showed just how many parts of the industry are already leaning into modern technology. It feels like the start of a shift, and it’s an exciting future to be part of.
Please keep an eye out for our follow-up technical blog on the paper titled “Automated extraction of geotechnical data from core photography for Sydney Metro West using artificial intelligence“.
