The CIO's big data challenge: Asking the right questions, connecting the dots
“The hardest thing about big data”, says David Thornewill, EVP, IT, Global Business Services at Deutsche Post DHL Group, “isn’t so much about the manipulation of data itself.” Rather, he narrows it down to two things: coming up with the right question, and sourcing the right data that can answer that question.
Thornewill, a UK native who has spent nearly 30 years in logistics and supply chain, has seen the industry go through a transformation with the surge of IT and the internet. For him, the applicability of the wealth of data available now hinges on the basic details.
“At the end of the day, machines now are so powerful that we have the equipment there to do the work. The challenge in sourcing data is making sure the data actually represents what it’s actually supposed to represent. Data hygiene, data grooming, making data mappable — that’s half the challenge, and that doesn’t necessarily require a whole bunch of high-power tools. It’s about applying the right drilling technique, which requires brain work,” states Thornewill.
It’s this mindset that’s driving DHL to increase its data analytics capabilities by bringing in data scientists to deal with and make sense of massive volumes of information. “All the software vendors would have you buy their tools and say you can’t do big data without them. But the reality is, you can do a lot of stuff with big data with some fairly straightforward tools,” Thornewill surmises.
Operating in the logistics industry which deals with managing a complex web of information networks and in turn amasses a great deal of data each day, DHL is no stranger to the opportunities that big data analytics presents.
“When you actually look at it, we’ve actually been doing ‘big data’ for years,” Thornewill declares, citing DHL’s efforts in routing applications and analysis of capacity planning as examples. “What we’re finding now is that better tools have emerged to be able to do these things,” he says, adding that in the last two years DHL has been significantly focusing on “the things that aren’t as obvious. Things for which we have data, and those data sets need combining with second or third data sets, and even data sets we buy in from third parties—be that weather applications or macroeconomic data.”
DHL has identified three areas of opportunity to leverage big data within: enhancing the customer experience, increasing operational efficiencies, and exploring new business models.
“Some of the things we’ve been working on include being able to predict when customers might defect. We want to be able to act upon that and engage the customer before they do defect, if you will, and be able to persuade them that we’re as good a company to stay with as any,” he says.
Real-time route optimization, crowd-based pickup and delivery, strategic network planning and operational capacity planning comprise some of DHL’s initiatives to increase efficiencies across the company’s several business divisions.
When it comes to shipments driven by e-commerce, for example, Thornewill informs that they pay close attention to such trends and activities on the Internet. “We’re doing much more advanced load planning in the freight business. We look at factors like the weather, peaks happening in Amazon, and see how these will impact our freight shipments two to three days later,” he says.
With DHL collecting a wealth of information from customer and logistical data as well as third party-driven social data, integrating all these data sets from which to derive insights is crucial. With each of DHL’s division—from express and forwarding to supply chain and logistics—presenting its own unique need for information, Thornewill says that a key goal of the company is “not necessarily to have a uniform view of the entire business—it would be crazy to try and match up container shipments with our parcel shipments—but to have a series of platforms that are purpose-built to match with a specific division’s needs. “
“What we’re seeing as we look across the company is that there are points of light where people are very advanced, and then there are areas which are perhaps more unmarked spots on the map. Our goal is to spread the goodness—to make people aware that there’s new technology out there that can help them. It doesn’t always have to be at the leading, bleeding edge—as I say, we can do a lot with the tools that we already have by asking the right questions and in the right place in the business,” he explains.
Robotics, the cloud, and security
With its pilot project back in 2014 involving autonomous delivery flights by parcelcopter, DHL is well-placed and keen on exploring new technologies that could potentially disrupt the logistics sector.
Thornewill compares it to the invention of the printing press — how the synchronous invention of various technologies led to Johannes Gutenberg’s moveable type printing press revolutionized print technology, Thornewill is hopeful that DHL’s efforts in exploring and experimenting in areas like the Internet of Things (IoT), Cloud, and robotics could “at some point perhaps create for us a whole new view of the logistics we see today. We’re discovering things in robotics, autonomous vehicles, leveraging and using the cloud because the economics just make it compelling...stitching all of these together is our core competence,” he states.
Beyond the advantages that new technologies provide, however, Thornewill is also cognizant of the “darker side” these developments may bring, especially when it comes to security.
“As CIO, I’m definitely spending more time on this. Probably five to seven years ago I’d only be devoting five to six percent on the issue of security, but now it’s more like 40 percent of the time and I see this only going up,” he reveals, adding that figures in a Cisco report revealed the logistics Industry has been promoted from sixth to fifth in the list of sectors most at risk from security threats. “Threats are growing day to day—and we’re finding that whether it’s state actors, criminal actors or terrorist actors who just want to do damage, attacks are becoming very well regimented,” Thornewill says.
Before unlocking the many benefits IoT could potentially bring, for instance, Thornewill asserts the need to take a closer look at its downsides as well. “The world we built up over the years wasn’t necessarily built for a highly mobile, hyper threatening environment. It was built in a far more benign state of the world in the 90s, and now we have to toughen it up. And with things like IoT, we really have to think before we deploy these things,” he says.
“It can be a boon to us—who doesn’t like the advantages of an Uber or an Amazon?—but there’s a little bit of comparison perhaps to nuclear power in the ‘50s. It’s going to be great; endless energy forever, but there’s a dark side there. In the information management side of things, there is something there and we have to be very concerned about it because the damage can be done quickly and from very remote places,” he concludes.