CFG talks to Ian Craig, VP of Product Management at Ontario, Canada-based software solutions provider Descartes, about
- how Phase 1 development on the Descartes Core™ Unit Load Device (ULD) Tracking solution went,
- what Bluetooth tracking enhancements are scheduled for Phase 2,
and how COVID-19 has illustrated the usefulness of automation and digitalization when it comes to companies’ abilities to react quickly and flexibly to unforeseen
CFG: Given that COVID-19 is on everyone’s lips currently, how can automated Bluetooth ULD tracking help readjust supply chains and increase the ability to adapt to a volatile
IC: Supply chains until now have been optimized and tightened to such an extent that if anything goes wrong, it tends to be expensive to deal with. The coming world was already emerging
around more flexible supply chains and being able to adapt – something that is obviously going to be hugely important going forward from here. How does that relate? Well, the more automated and
more networked tracking is, the less you have to make manual arrangements, and the quicker you can deal with changes. Despite the contraction in the industry, it is still probable that people
will be pushing hard to have automated tracking and the means to having a flexible supply chain.
CFG: Automated ULD tracking is no doubt all the more important currently with the number of passenger airlines looking to operate as temporary cargo freighters.
IC: Yes! Especially since many of these passenger aircraft will not necessarily run to schedules, but as charters. And there are currently initiatives with some airlines that we’re dealing
with to actually have tracking inside the aircraft as well. When we started Bluetooth (BLE) tracking, everybody was concerned about transmissions on planes, but over past two to three years, BLE
has become the norm and every airline allows it. It’s becoming more flexible and in-aircraft tracking opens opportunities, too, in addition to the usual tracking to and from facilities at
CFG: In-aircraft tracking?
IC: We are working on things like monitoring fire-resistant containers within flights and being able to alert the crew based on temperature or smoke, for example. So that’s one major
initiative in the aircraft, which we’re hoping to demonstrate as a production system by the end of April. At the moment, we’re doing calibration on a test container to ascertain where an item
inside is either on fire or not on fire. We’re calibrating it to say, if we have this density of smoke particles, what signal are we sending? The container is currently being tested on the
ground, measuring different amounts of smoke and temperatures to calibrate the tags inside the container, and these tags have to be somewhat fireproof themselves. They transmit by Bluetooth and
we can pick that up on a radio or another device or even a phone. This is work in progress, as the customers determine how many hours one of those FRCs will survive. The particular carrier in
question is looking at this initiative from an ultimate containment point of view, should a flammable object get into a container despite all the x-ray and safety processes.
CFG: Certainly, with Lithium batteries, the rise of ecommerce and the possibilities of undeclared dangerous cargo, the risks are endless, so containment is a necessary initiative, for
sure. Coming to Phase 1 of the Descartes Core™ Unit Load Device rollout. What learnings are you taking away and how will these change the technology and capabilities going
IC: Phase 1 has been a massive exercise. We’ve learned about cellular reliability and coverage, and this has pushed us towards new technology, what we call a mesh network, which works with
solar powered relays. These stand-alone relays which are just slightly larger than your hand, can be put on aprons and hardstands and all these sorts of places where you can’t get power and
therefore you can’t track. We’ve been asked about putting them on containers as well. So, we’re doing a lot of technology to improve the coverage and make things resilient to cellular
fluctuations. And we’re working on distance coverage. In a mesh, you can hop across multiple nodes before you get to a cellular gateway and the data is transmitted. You can do half a kilometer
with two or three of those devices, with the advantage being that because you know which node read your ULD, you can actually tell approximately where it is. We and our customers such as Air New
Zealand and Delta, for example, are quite excited about this, as there are currently areas that they can’t cover but would want to, so we’re talking about setting up an initial test installation
with them. A solar based mesh has been operational in Christchurch for nearly a month.
Another thing is, we started our BLE tracking initiative off on an 80% coverage case. Many players out there have expensive, top end data loggers, or tracking recorders, which generally have
to be taken home to base to retrieve the data, or they are cellular enabled and hellishly expensive. We decided to deal with an 80% solution that could answer the “where is my ULD?” question to
start with, and this led us to employ cheap, less sophisticated tags, but with a price point that makes it feasible to put them on every ULD, as opposed to just using them for specialized
shipments. We are now in sort of a half-way house, where our new tags can actually also track things like humidity, precise temperature and accelerometers to tell whether they’ve been shocked or
moved, but they’re still in that format of a relatively cheap tag that goes on everything. So, we’re eating into the 20% of the top at a different price bracket and with a lot more coverage. Our
tags are capable of data logging. We haven’t been doing much of it at the moment, but are heading that way, and this level of sophistication will come in at a reasonable price point and give
CFG: So, at the end of Phase 1, you have reasonably priced tags that are much more versatile than you’re currently using them for and provide an excellent basis for Phase 2. What does
this entail and what are the challenges?
IC: We’ve enabled the tracking of ULDs, and so the next thing is to stitch together what’s in or on a ULD and so that’s where Descartes and actually CORE have some good synergies because
Descartes have tracking systems at waybill level, driven by messaging. We can integrate the automated messaging from the Bluetooth network to make a match: “If I can see this tag – I know it’s
that ULD, but I also know what’s in it”. This gets us to consignment level tracking, which is the next challenge.
Phase 2 is tackling mesh technologies, resilient gateways, and our new tags, and incorporates what we’ve learned over the last 18 months. There are still some challenges. One in particular is
the brutal handling of pallets. It is very difficult to get a robust piece of technology onto a flat plate. So, we’ve been working on that. The physical challenge of a robust device, and then the
challenge of adequate coverage. Our network is not just for one party in a single warehouse scenario, it tracks everything around the airport, including ground service equipment and the like, and
relates it all together to give a much bigger picture than simply “I can track a ULD”: I know that I’ve got the right number of dollies, the right ULDs, at the right gate at the right
Our technology is important, but the next thing is using the intelligence. There’s a huge amount of information flowing through our system, and we can see patterns. After all, most things are
scheduled in the airline industry. We can see containers belonging to Air New Zealand coming down the road from the northern terminal at Hong Kong, and we know they’re going to a warehouse at the
southern end of Hong Kong International Airport. So, we build up an expectation and can say, for example, if it’s 4 p.m. on Tuesday and we’re not seeing Air New Zealand containers on that road,
then something’s wrong. Pattern recognition is the next step up, so that we look at exceptions rather than seeing everything that’s happening. We need to get more intelligent about what you
really need to know. Our network gateways are really quite powerful computers. We expect to learn patterns and then tell the network to behave differently in different places and only alert us if
something has gone wrong, or a parameter like temperature has gone out of range. Currently, we track on a time basis and we see it every five minutes or whatever timeframe we choose. Whereas,
we’re headed to “don’t tell me anything unless something has changed”. I’m looking forward to exploiting all that computing power put down on the edge of the network, around airports and things
CFG: So, it’s basically about business intelligence, big data, then you can do predictive analysis: if something’s moving then it should be here at a certain point if it’s going the
right direction. Are you working together with any start-ups or universities to make sense of all this big data or is that all in house?
IC: Actually, we’ve always had a relationship with a group called Callaghan in New Zealand which is government funded to a degree, and we get R&D grants from the government. The
scientists at Callaghan look at questions about the optimum number of, say, Bluetooth readers to be able to determine where something is to fine degree. They undertake a number of research
projects on our behalf, the results determine what we develop. One project is called “localization” and about being able to tell how far something is away. Bluetooth is almost too good, because
it can see for 200 or 300 meters. You can know what’s at the airport, but you still have to know where it is. So, there’s complex research on triangulation to get accuracy down to something like
6m which is useful on an airport, but that still isn’t on a rack in a warehouse. It’s still not that precise. Those scientists have developed a set of algorithms that we test out in the cloud and
we’re then going to push those algorithms down to the computers, the readers on the edge, for two reasons: 1) the data will be more responsive down there, and 2) it is all very nice having masses
of data but transmitting all of that over cellular is an expensive proposition. So, we’re optimizing what the network tells us based on these sorts of rules that we push down to the
Phase 2 is also concerned with a more open network. We have an “Open Network initiative” and because the readers are very intelligent in computing terms, we’re going to open it up to third
parties. Naturally this will have a few rules regarding identification, for example, but the objective is to allow customers to choose a variety of different tag technologies. This is important
since one thing that ground handlers fear is 20 different companies coming along and saying “Well, I need my reader on the wall”. There is a sort of drive amongst the people doing these things to
try and keep the network open. At the end of the day, we want the data, and other people can connect and provide useful pieces of data that we can send to their customers.
We have over 1000 readers out there at hundreds of airports. So, when a new airline joins, there’s a very high probability that we already have coverage in the network. It’s a very shared
network, so it doesn’t mean that for every airline, they have to install new equipment. Plus, the system itself has APIs which interface to our customers’ own computer systems too, allowing them
to provide their own view of the world through their tools. We do provide graphical interfaces, but any major operation has its own system and is looking to get enhanced, reliable, and excellent
data without all the manual inputs. The interfaces are just as important as the user tools at the front end of our service. Many of our customers see the value of an automated end-to-end process
that this solution offers, and that it is not just the inventory management of ULDs.
CFG: We wanted to meet at the World Cargo Symposium 10-12MAR20 originally. What, if you’ll pardon the pun, would have been your CORE message there?
IC: “Openness is important!” And we need to recognize that we are on a technological journey and it never stands still. So, for example, when people interested in BLE trackers ask, “How long
will the battery last?”, I tell them I could make it last eight years, but since technology changes so fast, it’s not worth doing that as opposed to keeping the price reasonable and totally
swapping out tag technology every three years. Originally there were things that we couldn’t do three years ago at a reasonable price, however now most of the movement and humidity sensors have
come down so much in price that it’s reasonable to put them on every device as opposed to having specialized devices. So that’s what’s driving the change. That, and getting more comfortable with
Bluetooth because it’s driven not just by our industry: everything has Bluetooth in it! All kinds of consumer devices, and therefore, because of this widespread use, we can rely on keeping the
price point reasonable. So, I think openness, ever advancing technology and lastly, integration with all the other major industry players. I think people are pushing for consignment level tagging
now, and we will get to the point eventually, where tag prices come down a level where you could throw one on every package or every consignment, but we’re not there just yet.
CFG: Thank you, Ian Craig!
We always welcome your comments to our articles. However, we can only publish them when the sender name is authentic.