Smart Traffic Solutions for Connected Cities in Australia/New Zealand

Randi: It's quite a challenge, because there are so many partners in play here, and more often than not, the people that put together these solutions, as the question was said, are not the people that actually end up implementing it, and it's difficult for that transfer of knowledge that, in an ideal world, would happen. Not everyone's a network engineer. Not everyone's an RF specialist. 

So, one of the things that we've tried to do as an organization, with Digi as well, is try and take the burden of these deployments off the feet on the ground. There are a lot of complexities with cellular routers, authentication, getting on the network, making sure you've got the right IP address, making sure you've got the right hardening profile, as per the specs of that particular network and architecture. Some ports should be disabled, some ports should be enabled, some features should be encrypted, etc., etc.

Our solution to that is actually taking on that burden, and doing the config in-house, doing the provisioning in-house. We receive SIM cards from the departments. We receive config requirements and application settings from the feet on the ground, saying, "Hey, guys, I've got XYZ CCTVs," or VMSs, or FDSs, "going out on site, and this is our config. These are the site numbers." 

We do the site-specific config, get it all working, make sure it's commissioned on the network, in our provisioning facilities, before sending out to the contractors to effectively take it out on-site. And it is, for all intents and purposes, a plug-and-play device by the time it reaches the contractors that do the work, do the hard yards.

Steve: We've transitioned more and more to 5G in our transportation solutions. Many of our products now are 5G. Of course, something that supports 5G also supports 4G. They never lose that, especially because of non-standalone, the 4G is actually needed. And even a number of them support 3G, as well. So, they cover the stack of technologies. More and more, of course, 5G is where systems are headed. Around the corner is something called 5G RedCap. If you haven't heard, it stands for reduced capacity, where 5G is designed in these particular situations to address lower-capacity, lower-power applications. The 5G we're all familiar with today is for high-speed, high-capacity systems. That's evolving too.

Randi: If I can just add to that, I will point out, Digi IX40, I think, will be a favorite in this region. It has just recently been released. It's still part of Digi's IX line, which is the industrial side of things. It's a nice little device, with four Ethernet ports and SFP. So, with the use of VDSL SFP, for example, you could pretty much knock off every single connection option that's available in Australia. You've got fiber, you've got VDSL, you've got Ethernet. So, if there's, like, an NBN handoff via an Ethernet connection, you've got that covered, as well as cellular. And with the Linux capabilities that the Digi solutions sort of show off, you can pretty much prioritize up to five or six different connections, if you need it to, over the course of fdIP tables and weighted connections.

Steve: I want to add one more thing. So, I don't know if you know that there's an incredible number of LTE devices worldwide, just an incredible number. 5G devices aren't even close to that. So, LTE is going to be around for quite some time.

Steve: Sure. As edge computing has caught on, more and more systems are involved in crunching lots of data, coming up with answers and solutions at the edge. How do you do that? Do you have to have a compute capability on the device? Digi, forever, has had Python as part of its offering, and now we also provide the ability to load containers, run containers on Digi devices. So, that's a whole 'nother new way to run, as a virtual operating system within a Digi router, most any application you can think of.

Randi: Yeah. And Digi IX40, with the quad-core processor, gives you all the processing power and capability you need to run any sort of custom scripts or custom containers that are required for the application.

Corey: Obviously, big deployments, such as New York City, throughout Australia, New Zealand as well, are going to require some in-depth analysis and work with the current infrastructure.

Steve: No. You know, SCATS originally was serial-based, but over time, it's now become IP-based. And so, SCATS, SCATS messages — the formats are the same — evolved over time, as all protocols do. But they're carried through all these networks using an IP-based communication system. So, it's very transparent. Oakland County, in Michigan, has been running SCATS for 30 years plus, and shifted to cellular about 10 years ago. And there are other systems, too, that I know of in the U.S., that are all running SCATS. So, it's quite universal now.

Steve: It's a great question, because, too often, the answer is thought that it's not possible, right? How do you protect a private network if you're connecting into a public cloud? But if you think back to when IPSec VPNs were invented, I can imagine many people wouldn't dare put their system traffic through a public network, right? But now everyone accepts IPSec VPNs as secure. Similar to a cloud, right, they have VPCs. They're virtual private clouds. They're quite secure. There's no public access to these private clouds, and you can access a public cloud, via an IPSec VPN, into the VPC of that instance, and stay off the public Internet completely. So, no public IP addresses; it's secure, and isolated from would-be hackers.

Randi: Yeah. And just to emphasize on that, with this architecture that Digi has put together, all the edge assets are not actually connecting to AWS. They're just connecting into the TMC, via a proxy server. So, the units themselves are told to send their Digi Remote Manager® information to a proxy server, and it's just that proxy server, going through that VPN connection, that sends management traffic across to Digi Remote Manager. The actual OT traffic, of edge assets — you know, your CCTV footage, your VMS messages — none of that touches Digi RM, and vice versa. So, it's just that two-way traffic, Digi RM to the edge assets, via the proxy, for any sort of config and firmware upgrades, auditing, compliance checks, that sort of thing, and the devices, the edge asset, the cellular connections, going out to Digi Remote Manager, to send health stats and things like that. But any sensitive information stays within your network.

Steve: The way it's protected, and I believe it's actually an approved method in GDPR, but there's a one-way hash. It's a highly mathematical algorithm, that takes the MAC address and transforms it into a unique number. The way it's structured, there is absolutely no way the unique number could be transferred back into the original MAC address. You can't do it. And it's just fine that way, because these systems who need a MAC address are looking just for a unique number, and the algorithm preserves that. The unique number is unique. A different MAC address running through their algorithm won't produce the same unique number. So, it works just fine. Everyone's identities are protected, and these traffic systems can use the data for traffic analysis.

Steve: As I mentioned earlier, QoS — quality of service, an Internet standard — is the best way to do it. It's been around for decades, and all you do is mark those packets with high priority in that DSCF field, and the whole system, not just the Digi router, but every router downstream, treats that packet with priority, putting it in the front of queues wherever its destination route is set up to be. So, it's proven. It's not used enough. It really needs to be turned on, and it's a good way to make sure that things such as Connected Vehicle, or the traffic controller packets and communications stay at the top of the performance curve.

Corey: You make it sound easy, Steve.

Steve: No, I don't know how to do it. Others do, like the engineers in Digi Professional Services.

Corey: That's a good note on support. Obviously, we have a great team, and a lot of different people from support to the engineers building the solutions themselves.

Randi: Yeah, absolutely. SAPPLY serves as a conduit for Digi, and Digi has a phenomenal base of engineers for their support. So, it's not your typical, you know, send through your problem and we'll escalate it to someone in 48 hours' time. More or less, the person that receives that ticket will be the one that sees it through to the finish. We ensure that any ticket is addressed, and if it does need to be escalated, we've got local resources where we can escalate it. So, Digi's resources in Brisbane are technically R&D and engineering. They're responsible for building all the wonderful pieces of hardware that come out of Digi, so they're not typically part of the support escalation path. But, you know, we always kind of leverage it if and when required. If it's something really funky going on, we can always ask for a bit of help.

Randi: Yeah, absolutely. I mean, if it's something that's never been done before, proof of concept is a no-brainer. So, typically, in the proof of concepts that we've done in the past, we'd start off with just one or two units on the test bench, prove the config, prove the functionality, test anything else that you might want to have a look at, such as disaster recovery, that sort of thing. Then we’d take it into a slightly smaller sample in the field, maybe 20 to 50 units, test that out for two or three months, make sure there's nothing else that's red-flagged, something that's been missed during that test bench phase. 

And during that field trial is really when we sink our teeth into the actual logistics of the program. So, let's say we're talking about a deployment of 5000 units. That needs to be planned. There's quite a bit of intricacy in terms of scheduling hardware, scheduling provisioning, scheduling SIM cards. All of those pieces, we try and be involved in, and make sure that we provide a cohesive, turnkey solution to the end client, with the cooperation of any other stakeholders that are involved — be it contractors with the work package, maintenance teams, delivery teams, cyber security, etc.

Steve: Yeah. I think, just looking at New York City, to think that a mission-critical system such as that would run on a commercial cellular network 10 years ago, probably people would have thought you're nuts. But what's happened is the evolution of cellular. It has a lot to do with what the demands are, that folks need these systems for, but also 3GPP, who's designed the system and specifies it on a global basis. 

Ten years ago, they put into the network architecture the ability to do priority and preemption. But that's a specification. It was a challenge, right, in its design and deployment. It's now prevalent in all major cellular networks. You can pay a little more, but you get priority and preemption. And places like New York, you can imagine. It's quite frequent, when the network becomes loaded in certain regions, the traffic system stays up just fine, because every traffic cabinet gets priority over typical cellular users in the network. So, it works great. That's why cellular is on a climb rate to take over a lot of networks.

Randi: And just to speak to that, in terms of the New York City-specific deployment, I think what's fantastic is the dual cellular module. It's not just dual-SIM. It's dual cellular modules. So, you've got two carriers constantly dialed in. And so, if one carrier decides to unplug a core switch, for example, there's automatic failover onto that other carrier. So, I think that lends itself to a lot of robustness that could be leveraged.

Randi: Yeah, I think so. Running fiber is not cheap. I mean, you’ve got to dig up the road, you have to plan for that. There has to be capacity in the conduit. And sometimes you just don't have that luxury. So, with cellular becoming more and more prevalent, and with society becoming more and more dependent on it — and with 5G just around the corner, I mean it's already being deployed. 4G's is at the peak of its deployment. I don't see it going away anytime soon. 

Steve: Yeah, and just one important aspect to that, because, in our discussions with various cities who have fiber deployments predominantly, when there's a fiber cut, it's a fire drill, right? All hands on the site, because the corridor is down. And it's very costly and disruptive to people, and to the system, and to the users of the traffic system. So, fiber's great. Put a cellular backup in there. Just in case, so when there is a break, then no one has to run and get the trucks and go fix that fiber, right? The system's up. You’ve got some time. You can fix it over time. It just makes for better system operations.

Steve: Sure. It's going back a few years, but early on there's a specification. How do you actually design something to that? There's a lot of correspondence, and working with engineers at Digi, and engineers at DOT, to find the right solution. And then, as it evolves, it's a never-ending process. It was definitely collaborative, and we've all benefited, as Digi’s equipment and technology has improved. It has become tailored now for traffic systems because of that huge network.

Corey: I think that's something that's so important. Steve and Randi both mentioned the feet on the ground and the local experts as well. I think it's one thing to have the engineers, and the sort of brilliant mind-in-the-sky type of roles. But you really need to have the local understanding, whether it's in Brisbane, Sydney, New York City, wherever, to have that local support. That's something that Randi touched on as well, to have those local industry experts, as well as the people who are there to answer the everyday questions.

Randi: Regarding the feet on the ground versus the engineer-level knowledge base, it's impossible to transfer everything from the person that's designed a solution through to the person that's in front of a cabinet, trying to plug everything in, and get it all working. But I do think education is a big piece of that. Obviously, you have to be realistic and practical with education, and what the audience is. You don't need to be teaching the full TCP/IP stack to someone that's about to commission a router, but just general principles of what needs to be done, how things need to look. Simple things as well, like using a MIMO antenna in 4G rather, than a SISO. It's incredible how much of a big difference minor details like that make to the success of a deployment. And the better educated everyone is, the more successful these things tend to be.

Steve: Yes. We work with carriers. We've partnered with carriers. It's, of course, key to the solutions of cellar, and it's also evolving, and much appreciated. So, generally, I don't work too much outside of North America, but it's a key part of our solutions, and our partnership, working with carriers, and as their networks evolve, we evolve with them, to deploy these solutions for our customers. So, we all benefit from that.

Randi: Yeah, even from a local point of view, it's kind of a no-brainer, really. Like, the cellular routers won't work without a carrier in place. But from a dynamic point of view, you need to almost have this three-point system, where you've got the client that's effectively going to be taking the service from the carrier, and then you’ve got the carrier and the vendor as well. And if any one of those parties is out of line, out of whack, it's just not going to be cohesive. So, it's not just about the client and the carrier having conversations, and the manufacturer, where it does its thing on the side, and vice versa. Everyone kind of needs to be on the same page, and Digi does that quite well, in terms of staying in touch with the latest changes that are happening with local carriers, and any patches that need to be created. Digi's Brisbane office releases a firmware update every quarter, with any major releases, any security upgrades. Any severe security vulnerabilities that are identified during that quarter will be released in a patch. So, any network changes from the carriers' point of view needs to be considered when these upgrades are being built. And so, kudos to Digi on that front, in terms of keeping up with everything that's going on.

Steve: It's definitely a partnership.

Corey: Yeah. Right down to the language and the locals, even. I get messages from major carriers if any brand guidelines shift or anything. So, I know that's a big thing we really pride ourselves on, is playing well with others, and making sure that we're in good step with the carriers, as a key part of that equation.