Tracking, Locating and Positioning - What is the difference?

In this episode, Talking IoT host Steve Carr speaks to Jimmy Wind - CEO of Versa, and Warwick Taws, CTO of WiTTRA Networks AB - about unpicking why these terms are commonly misused interchangeably, and the challenges that come with it.

Transcript

Steve Carr (SC): Hi, my name is Steve Carr, your host on talking IoT. 

Have you ever wondered what the difference is between tracking, locating and positioning within IoT? The unfortunate truth is that in many organisations, they interchange these very different terms when defining their project requirements assuming that they are one and the same. 

The distinction matters and can prove very costly, delay deployments, or even result in project failure. Today we will be discussing these terms, which are wrongly and commonly interchanged. 

I'm very pleased to have with us today. Two guests, Jimmy Wind. Jimmy is the CEO of Versa who are focused on driving smart tracking solutions to optimise logistics and Warwick Taws, Warwick is the CTO of a company called WiTTRA who looks to answer the question, where is my stuff and what is it doing? 

We'll start with a brief introduction from our guests. Jimmy, why don't you go first and tell us a little about yourself and your organisation? 

Jimmy Wind (JW): Alright, thanks for having us on the show. My name is Jimmy Winds. I'm from Versa, Since 2016. Originally we focused on improving the baggage service for airline tracking their check-in bags. And then we ventured out into a more broader range of solutions, to track assets and shipments in logistics to help optimise logistics for different kinds of use cases. 

(SC): And Warwick, a brief introduction to yourself and to WiTTRA. 

Warwick Taws (WT): Thanks, Steve. Warwick Taws is my name. CTO of WiTTRA, I've been there for about eight and a half years now. We focus on, answering the question, as you said, where’s my stuff, and what is it doing? And that naturally needs two different kinds of data. One is sensor data to provide context for an asset tracking tag – an asset management tag - and the other is positioning data. Hence we get into the definition of positioning versus tracking. 

(SC): Thanks guys, and welcome. So Warwick, what do you see as the fundamental difference between tracking, locating and positioning? 

(WT): Yeah, that's a good question and I think often these terms are used interchangeably. From our perspective when we talk about the positioning of an object or an asset, we're referring to where it is – physically - in that building at a given point in time. Tracking is more referring to, if you look at the, let's say a map and you want to see the path of something that's moving, the path that it's taking or has taken like a trail of breadcrumbs. We would refer to that more as tracking, so being able to see where the object is going during its journey.  Their location/position, I think that's where they are used interchangeably, and for us, it can be a subtle difference. But you know, many people talk about a real-time location system and again, that’s then getting back into tracking. So the real-time element is where that comes in. Location, for me, it's fairly similar to positioning. 

(SC): Jimmy, from what you said at the start of your introduction, Versa seems to be very much a tracking technology. Is it the one technology that you use, or do you use multiple devices within your deployment in order to track? 

(JW): No we use multiple devices, Steve, good question. So actually we were fully focused on tracking, shipments, so tracking a shipment, parcel, which was transported from A to B. Using multiple modes of transport and really enabling the shipper, the freight-forwarder and also the end customer to know where a shipment is, what the expected estimated time of arrival was. 

And after the fact enable a breadcrumb trail to see where the package was and if something went wrong, then to pinpoint where it went wrong. Let's say, for example, a shipment of medicine. If the temperature was breached that she could see exactly where the temperature deviation took place. This would enable discussions with insurance companies, for example, to help solve that question. 

(SC): So that's really interesting. One of the things that you're doing in terms of tracking, you're following on from Warwick's initial comment, where they're actually sensing as well. So you are actually sensing the devices as well as tracking the devices because many devices on the market today are just pure trackers, right? From what you've just said, your solution actually has some sensing capability associated with it.

(JW): Yeah. So we have sensing capabilities in the sense that we can sense conditions like temperature, humidity and light. But we also sense on which mode of transportation the tracker is which helps us optimise the performance. Because we could say if the trackers are on the van, you need a more frequent update of the location because it's approaching its destination. Whereas if the devices are on an ocean freighter, you don't need a quarterly update. Because it would drain the battery, it doesn't make any sense on a three-week journey to have an update every 15 minutes. So we know which mode the tracker is, and the shipment, and then we dynamically adjust the tracker. This started with the initial product that we developed for aircraft transportation, and in the tracker, we incorporated a receiver which receives the aeroplane signal. So we know on which aeroplane the tracker is, which provides very relevant information. But it also enabled us to switch off the device during the flight to save battery and to be compliant with all the Aviation regulations globally. 

(SC): I love the fact that this is really a smart solution for logistics, especially with the FAA   approvals you managed to get, and also the Marine side. We talked at the start about the complication, or the interchangeability of these terms, one of the other clear aspects, which I think confuses the marketplace somewhat is the difference between what is real-time location and what isn't. And also the indoor and outdoor scenarios. Real-time location appears to be the holy grail that, people, try to pursue where actually that isn't a requirement. What are your thoughts on that subject matter? 

(WT): That's a good question, Steve, and this often comes up in discussions with customers. I think at the end of the day, you need to understand the use case and more particularly what are the problems or the problem, that the customer is trying to solve. And there can be more than one of course. But of course, when you talk about tracking a device, if a power source is available on that device, you can power your own asset tag. When you're in a battery-powered scenario, which is the vast majority of use cases that we are involved in. And then you always, always come back to this equation of, update rate - What do you need for update rate? Versus battery life. You get backed into a corner because the customer is, oh yeah, I need a one-second update rate. Yeah, because that's what I get on my GPS on my phone. But when you start analysing the business case, very often it's not like that at all. And you can do, you know, as Jimmy just outlined, you can do smart things in the asset tag itself, context awareness to optimise how often you actually need to get a position update. There are two aspects, taking the position generally involves using energy and then transmitting that information off the tag to infrastructure takes energy as well. So, it can be a  complicated question, but really it's what's the use case? What are you trying to solve here? And then that really drives the design of the system.

(SC): We've seen instances where customers are looking for complete tracking with inside an indoor type of environment, which is very different from tracking from an external type of environment. You don't have necessarily a line of sight and you don't have a GPS. In those types of scenarios, what would you recommend or how would you help them in terms of scoping that project? For example, if somebody said they wanted to track the Forklift running around a warehouse. How would you actually go about doing that? What could they expect to see? Because a breadcrumb doesn't really work in an effective manner. 

(WT): Yeah. Again, I guess it gets back to this energy equation. The forklift has an energy source. So in that sense, maybe you don't have to worry so much about a battery because you can then provide external power for your asset tag. Then it's then it gets back to the infrastructure that's needed. 

To track that, If you were talking about a use case where you have a handful of forklifts, in a very large warehouse, it doesn't make sense for an enormously expensive infrastructure to track a handful of forklifts. So, you know, you have to look around at the options you have for that particular offering. It is very suited to these kinds of use cases because we have a low-density mesh infrastructure. This means that we have very few mesh anchor points and we can still position reasonably accurately. With our infrastructure, if you try doing it, for example, over a wifi network or even a BLE network, you need a pretty high density of beacons or anchor points in that setup. We're typically all about solving these kinds of problems on a site. By that, I mean, it's got a defined boundary geographically - then we can cover that site with our mesh whether it's indoors or outdoors or underground or up in a high-rise building. We have a 3D positioning engine and it just takes its positioning references from the mesh anchor points, and because we're using sub-gigahertz frequencies, and we have a fairly sophisticated way of measuring the position of the asset tags we can use this load density network. With relatively few anchor points. 

(SC): Makes a lot of sense. Just Jimmy, coming back to yourself.  Your solution appears to be much more externally orientated. Is that correct? Or do you track externally and internally, 

(JW): so originally, as I said, previously, we focused on external tracking and tracking from A to B. On a global level using the aeroplane as the main means of transport. But because of the fact that we started with tracking baggage, where it goes wrong is predominantly an indoor type of area, we did have a pretty good focus on indoor as well right from the start. And to make sure that we enabled ground handlers to track things like bags, in a dense baggage hall, we used the combination of generic tracking using things like GSM cell tower sniffing and wifi sniffing together to get a pretty good idea of where the bag would be, but then compliment this with BLE tracking, and we developed a smartphone to point the baggage handler towards the bag. A combination of generic tracking, based on the tracking capabilities in the tracker, but we complement that with other ways to do more proximity-based tracking indoors. 

(SC): One of the key challenges I guess, comes down to the range, and location accuracy once you've tracked the device. The device is somewhere. You now want to locate that device. What positioning accuracy can you achieve? 

(JW): You know, so that's following the app, basically. So the baggage handler would see the bag on his screen and he would see arrows pointing him, in an augmented reality kind of way, towards the bag, and see the remaining distance to the bag. 

(SC): Okay. It's clearly one of the big challenges that people actually have. So when people are actually looking to define that project, what do you see? And it's really Warwick, to yourself first of all, what do you see as to core perceived needs versus actual requirements that they should be looking at? What are the key external factors that need consideration when they're actually defining a solution? 

(WT): You know, defining the requirements properly is always a challenge, and we see that when we come and talk to a customer, you can get one view from the operations manager or the site manager, but you go to talk to the people on the floor who are actually doing the work, and they have, sometimes quite a different view of what the needs are and what the problem to be. So I think defining the requirements with as many stakeholders as you can get hold of is important to understand the problem properly. You have to look at the context of the site. Often you can go to a site and start with one problem or a set of problems to solve, and then you do a proof of concept - we call it a proof of value - for a trial like a month or two months. And at the end of that, there are five additional requirements that have been added in by the customer because they've seen the possibility to solve other problems with the system. So, it's important to be able to offer a technology solution, that is scalable and expandable. Because what we've seen a lot in the industry is IoT vendors who are peddling one particular flavour, or one type of technology, and walking into customers and saying, hey, this is going to solve all your problems. But then it doesn’t. 

(SC): So that goes back to the unification argument, and I think in this instance, both Versa and WiTTRA looking at your product sets and how those product sets are put together, use multiple standards and multiple protocols in order to achieve tracking and positioning. One of the key challenges that some clients have is they want to be able to track. When we talk A to B and Jimmy, you talked about your tracking A to B from a flight perspective. However many of them, when it lands in a particular building, might want to go down to what we defined as location accuracy. A Point on a shelf or in a room, which obviously is not covered by the Versa technology. How would you integrate or how can you integrate your technology with other technology? Maybe to add a little bit more colour around that statement, if you take the WiTTRA solution, could you, in theory, connect to a WiTTRA device so that when the WiTTRA device leaves the campus, the Versa device could take over? 

(JW): Yeah, we could so that's what we basically developed. What we came across were questions from customers that said, okay, so your tracker is a great product. But putting a tracker everywhere would cost us a fortune, so isn't there a way multiple boxes in the shipment without having to add lots of trackers? So that's when we looked at ways to do that. And what we did is made our tracker into a BLE gateway. So any type, BLE, tag or label. So we are working with multiple vendors to develop these types of labels depending on the required battery lifetime. Whether it's an outdoor or indoor use case, we can select the right label or tag, identify that based on the BLE signature as a friendly tag in our system and then any tag of that particular brand that it's picked up by one of our trackers will collect the data and send it to the platform. Then we expose that with the API, which we also use for our tracker. 

And to extend that we also created a version of the tracker which is not battery-powered but which is connected to power or power over ethernet. So that also, we can provide continuity by also being able to collect the data in stationary situations like warehouses or retail locations where we have use cases. 

So it's creating an open set of technology so that we can connect. We don't think that, I think Warwick already pointed it out, there's not one Swiss army knife type tracking that you can combine. There are different types of hardware, they need in different situations for different customer requirements. 

(SC): You just said a really important thing. Many customers are looking for, that silver bullet, that they're looking for that technology, that solution that gives them everything in one deployment, and I think unifying - not just the various protocols and standards out there - but actually companies unifying technologies in order to cover it really is an important aspect of being successful within the space. 

In terms of external positioning, Jimmy,  I know you're using cellular, you're using LoRA, you're using a number of other technologies in there.  Are you using satellite? Is satellite something that you'll be considering with your technology? 

(JW): We talked about it. Quite a bit, but we haven't added it to the solution. Because in a terrestrial situation, GSM coverage is normally pretty good. And, you have to look at it from a coverage perspective, but also from a cost perspective. So we haven't found any use cases that have a big market potential, which require satellite. And adding satellite to a product obviously would increase the phone cost, and it would also increase the subscription fee. So, as I said, there's no Swiss army knife of trackers so we try to not be too broad on the scope of our product too much. 

(SC): Makes absolute sense. Just coming back to WiTTRA, and to Warwick. Warwick, we have spoken a bit about how you deploy the anchor points. What type of technology are you actually using to do your positioning?

(WT): Yeah, our mesh is built on the sub gigahertz. We use 6LowPAN so that gives us mesh. Having anchor points provides reference points, and reference locations, to be able to do a form of triangulation. So it's a little bit like the way GPS satellites work. 

We are measuring the time that the radio signal takes to travel. From an asset tag to an anchor point and back, and then we do that for several anchor points. So if you have more than three different anchor points, you can do a triangulation - it's actually called trilateration, but most people know triangulation calculation - which gives you a three-dimensional location on the site. And we then place that on a map of the site and a floor level. 

(SC): Can you get down to decimeter positioning? 

(WT): Okay. So the sub-gigahertz technology is not capable of resolving, too much better than a few meters. That's what happens when you have a narrow band signal, it's well-known as a physical constraint of radio systems in general. 

So we've taken the approach that if a customer needs sub-meters, sub-half meter accuracy, and some customers do need that, then we have a modular hardware concept and you can click on additional sensors and devices onto our basic asset tag. 

And one of our click ons is an ultra-wideband radio. UWB ultra-wideband is what it says. It uses an ultra-wide frequency spectrum to have high accuracy, high precision, locate or positioning. What we do is if there's an area of the site that needs this high accuracy positioning, then you click on UWB modules onto your anchor points in that zone, that area. And then you click on the UWB radio onto the tags, which correspondingly need this high accuracy. But you might not need this over the whole site. Typically you don’t, it's rooms or part of the warehouse, for example, where they need this accuracy. So you only need to spend this money on those areas that require that level of accuracy. The Several meter level of accuracy tends to be good enough for many use cases on other parts of the site. 

(SC): In theory, this is when you're getting down to decimeter, that would be, I guess defining UWB is RTLS. Real-time location solutions - because the signal latency’s not really there, it’s milliseconds. 

(WT): Yeah. So if, and I guess that's the other aspect of UWB is that because it can derive a position very, very quickly, because as you say these signals are nanosecond level signals on the air, A) it's very power efficient, and B) you can get a frequent position update because it doesn't spend much time on the air to derive a position. 

(SC): Jimmy just coming back on the subject of tracking, and positioning what we've just talked about in terms of real-time location systems. In terms of external tracking, I know you guys produce a breadcrumb. Is your device acting in real-time or is it acting at certain points in time? 

(JW): When we talk to companies that want to use our solution or are looking for a tracking solution, they want to have it all. So they want to have a live update and what Warwick was referring to, I often get asked as well. So can you show me a smartphone, like life, like the positioning update, like you see Uber car and the Uber app, and I want a very long battery life. That doesn't work together. So the way our system works is that you can set the interval. Which can be as short as a minute, or it can be it can be longer. 

We dynamically optimise the interval to avoid unnecessary battery drains. So our philosophy is to make tracking as easy as possible without too many manual activities. So that's why we don't have buttons on our tracking devices. They switch on based on motion, so if an object is stationary in a warehouse for a week our tracking device will go into a deep sleep and it won't do anything until it gets moving again. So that's what we do, but the tracker will automatically generate a location update based on the interval, which has happened until somebody changes it. And when we get an update we tried to do it as battery efficiently as possible. So if we're a lower network, we use it if the tracker sees enough cell towers and wifi access points, we do. And if all else fails, we invoke GPS. Which is consuming a lot of power. 

(SC): One of my reasons for asking it was really, there's a lot of confusion about bread-crumbing and what it looks like. So from an external perspective, if you're moving products or assets across multiple geographies, on a global basis, then there's a clear value in terms of having a visualisation of that breadcrumb. When you look at what's happening in terms of a warehouse-type environment, it doesn't appear to make a huge amount of sense, because all you're going to end up doing is having a lot of lines crossing each other at different points in time because a vehicle could have been, driving around the same racks all day. So it goes down to customers trying to understand exactly what they're trying to achieve, over what period, and how they do that.  

A final couple of questions. So, staying with yourself, Jimmy, first of all, what do you see as the next emerging IoT technologies for tracking and positioning?

(JW): I think wide band was already mentioned. But other than that, we try to focus on what's generally available because we want to provide solutions that don't require a lot of infrastructure investment. And we try to, because we were going to a global scale, make products which use well known, tried and tested technologies. So we're not too much focused on seeing new technologies, which would bring additional benefit, but, which are not tried and tested technologies

(SC): Warwick going across to yourself. Same question. You guys are focused much more on, on a campus orientation, so it might be a different answer. 

(WT): Yeah, for sure. UWB has been developing at a fast pace in the last couple of years. And, now it's been brought into the standards area. So there is a standard now for UWB on mobile phones. So that's been one area of fast development. And we see that will continue because of the benefits that it brings to high accuracy location. Has its downsides, very limited range, for example. It doesn't penetrate through walls very well. Antennas are very sensitive, but yeah. You can make the other area that we see is obviously we're continuing to develop our sub-gigahertz location technology. So providing a low-cost infrastructure with few anchor points, we'll continue to develop that. And alongside that part of that answer, we're making now making use of AI and machine learning in the network. To improve the reliability and precision of the positioning measurements.  

(SC): From my viewpoint and talking to both of you today, and in summary, customers need to scope better. The projects that they've got really understand significantly more, the external factors that they're looking to try and achieve. And actually move away from just assuming that everything is tracking its different requirements for different types of applications.  

Do you there a time when maybe there could be a WiTTRA/Versa type of collaboration and solution, allowing somebody to track internally/externally, it just appears to me that could be an awesome solution. Covering all eventualities. 

(WT): I can jump in there. Steve and say yes for sure. Because Bluetooth is such a globally deployed technology and it goes right across consumer, through, into industry, it's very relevant here and we also have a Bluetooth radio in our tag. We don't use it at the moment for locating or tracking purposes, but I could absolutely see a use case where we can make use of Versa’s BLE gateways and have our tags. Hop onto a Versa solution, and then hop off somewhere and go back into a WiTTRA network. So, absolutely. 

(JW): I see that to Steve because we often run into use cases where the level of accuracy required indoors on the campus, in a warehouse where they want to exactly know where a piece of clothing or book is, and then I think WiTTRA could provide additional value, which we don't provide with our current piece of hardware. 

(SC): It feels to me that not only are we trying to unify standards and protocols and deliver IoT backbones we're also looking here at how do we collaborate in order to really give that silver bullet solution to customers as we go forward. So gentlemen, thank you very much. I wish you and your organisations, every success and look forward to discussing your progress in the future. Thanks a lot, guys.