How does a multinational tracing its roots as far back as 1893 adapt to the digital age? Rail Express spoke with Thales digitisation expert Michael Livingstone about how the company is driving change through big data, machine learning, and artificial intelligence.
As the Australian director of Thales’s Guavus business unit, Michael Livingstone helps businesses discover how digital can bring them tangible benefits. Through workshops and business process modelling, Guavus evolves concepts into actionable steps, driving a digital transformation not only for Thales’s clients, but within Thales Group as a whole.
For what is a highly technical and at times mystifying subject, the gist of Thales’s digital push is quite simple: take more than 100 years of experience in industries like defence, transportation, aerospace and security, and enhance them with modern-day, data-driven analysis.
“Thales is both the provider, and potentially the customer, for each of our optimisation programs,” Livingstone tells Rail Express.
Once a Silicon Valley startup, Guavus over the last decade became a pioneer in real-time big data processing and analytics, and was acquired by Thales Group for more than US$200 million in 2017. At the time Thales CEO Patrice Caine said, “The application to Thales’s core businesses of Guavus’s technologies and expertise in big data analytics will strengthen our ability to support the digital transformation of our customers, whether in aeronautics, space, rail signaling, defense or security.”
That mentality is consistent with what Livingstone and his team are doing today in the Australian market.
“We’re looking at how we take our long institutional memory and apply that to this new digital footprint,” Livingstone says. “And for our clients, that means looking at new and emerging opportunities in their industries, too.”
Guavus was one of a series of key acquisitions made after Thales set out on its digital journey more than four years ago. Other recent additions include encryption specialist Vormetric, driver advisory system provider Cubris, and digital security firm Gemalto. Livingstone says opportunities for digital change both within and outside of Thales may lie in core processes, customers, suppliers, employees, or assets, but he says all are built from a strong understanding of current business practices.
“Historically there’s simply been an overwhelming amount of data for people to consider,” Livingstone explains. “In fact, in the past there’s been so much data there was very little motivation to collect more.
“There’s no point collecting real time data to analyse for an outcome of control, if that’s going to take three days. You need to be able to analyse the data, make the decision, and exercise the control in a time factor that is effective.
“That’s where the advancement in computer capability comes in, and the development of those analytics platforms themselves. Today, data generated in real time can be analysed in real time. That’s something that couldn’t have happened in the past.”
This combination of advancements in hardware, software, and the emerging fields of machine learning and artificial intelligence, are the core drivers of digital change.
Livingstone says there are plenty of opportunities for the rail sector to apply these sorts of technologies. “Digitisation is helping get more capacity on rail infrastructure,” he says.
“Analytics and smarter signalling can create an increase in capacity, better scheduling, and improved availability, all the while maintaining safety requirements. Then there’s predictive maintenance for both rollingstock and track, which can help an operator become significantly more efficient, and achieve a significant rise in network availability despite reductions to maintenance costs.”
Thales has success working with clients to enhance predictive maintenance through digital applications.
“The projects we’ve seen to date have seen some very significant benefits in terms of availability and overall cost. And we continue to work with our partners on the best outcome, because any analytics platform can be refined, more data can be captured, and layers of context can be built to give a greater degree of accuracy in terms of predictions.”
Smart Cities: Big data in action
A key digitisation concept for public transport is how rail can best fit into the concept of a ‘Smart City’.
“There’s a lot of focus at the moment on Smart Cities,” Livingstone says. “It’s a broad topic, and there are different applications; it can range from a rubbish bin in a park telling the city when it’s full, to tracking your pet.”
From the perspective of public transport, Livingstone says a Smart City helps its residents become more engaged with the city itself. It helps them take better advantage of public transport, and drives more people towards it. On top of this, a Smart City better understands traffic flows, and the flow of people, and can minimise jams through optimisation of things like traffic lights.
“All of that is a data and analytics problem,” Livingstone explains. “We want to be able to see how things work in real time, and then what we can do to improve on those processes, how we can get predictive on those processes.
“The transportation industry is moving more towards autonomy and the use of analytics in transport and logistics, and like in other sectors, the end user is expecting more services to be attuned to them as an individual: ‘I want to understand where my train is, why it’s late, when it’s going to be here, and what are the alternatives.’”
One key area Livingstone believes rail transport can see immediate benefits from digital development is security and safety.
Traditional CCTV systems, he says, require more eyeballs as they grow, and still rely on the effectiveness of a human observer to catch unusual events, such as a person behaving in a suspicious manner, or experiencing a medical emergency.
“Once you consider the digital progress made in this field you can ask, ‘do we need to capture footage at all?’ The answer is probably no.
“If you can instead look at people moving around a train station as objects, then by using machine learning you can understand how people flow through that space. When the system then sees something unusual – for example an object enters the area and remains stationary for an unusual period of time – then you can engage CCTV and potentially security staff to focus on that area.
“In this way we can use data analytics and technologies that don’t identify individuals, to observe an environment objectively, and then the technology that can actually identify people only needs to be used when an abnormality occurs.”
The large-scale analysis of a flow of individuals through a train station of course has added benefits, beyond just security and safety.
“We can use this knowledge to get the most out of our assets. In ground transportation we generally have a greater influx of passengers over time, and a limited capacity,” Livingstone says.
“Using real time analysis, we can engage with commuters through visual and audio alerts and their devices to better optimise this flow, as it happens.”