Q&A: Telematics and fleet tracking applications

In September, Navigil, an original design manufacturer of programmable GNSS tracking devices for intelligent track-and-trace applications, released theTD230 OEM Tracker. The TD230 is a white-label wireless tracking device that consumes a tiny amount of power.

Navigil aims to provide it as an enabler of track-and-trace applications including personal safety, fleet management, vehicle black boxes, on-board diagnostics and insurance telematics. Navigil CEO Matti Räty talks to TU’s Susan Kuchinskashow embedded processing capability could decrease the cost and reduce the time to market of innovations in tracking applications.

Track-and-trace is a well-developed market. What's your competitive edge?

There are three key things that set us apart: in-built intelligence, re-brandability and power consumption. Making the device intelligent enables us to minimize the amount of data being transmitted, and lets customers choose the granularity of the data. Re-brandability means the device can be made the way the customer wants, whether software or firmware. Power consumption is a big issue; we have low power consumption. In so many places, when you don't have constant power available, that's an issue.

What applications can the TD230 OEM Tracker enable?

It is extremely well positioned for UBI because of the intelligence we built into it. Companies are talking about ‘big data’; one fleet tracking company has over one million vehicles reporting to their servers. If these devices report every second where they are, the amount of data becomes quite large. With our product, you can define how it works. We've got close to 1,000 parameters you can adjust individually, if you wish.

For example, you could set it to have to send a report when any one of these things happens: every two minutes, every X meters, or when there's a heading change. For the insurance markets particularly, you can set the granularity of the data based on the event. For example, an insurance company wants to know if there was harsh acceleration or deceleration. The accelerator reporting doesn't happen at all unless we get 2 Gs of acceleration. In that event, you will get data for 10 seconds before and 20 seconds after. (For more on UBI and ‘big data’, see Industry insight: Fleet telematics.)

How does Navigil handle identifying what data is valuable?

Insurance applications can be updated over the air very easily. Imagine an insurance customer wants to go to a new market. At first, we'll take every customer and get highly granular data for two months, and then turn it off so only they get data around bumps, swerving, etc. So, there is high granularity for underwriting purposes and data mining, and then switch it off. (For more on insurance applications, see Benchmarking with fleet telematics and Managing driver behavior with fleet telematics.)

Your new devices are pre-programmed to support commonly used features for fastest time to market. What are some of the most common features? And what are some more advanced use cases you support?

Embedded intelligence is something you can easily build on, for example, automatic driver rating. We can do that inside the box, you don't need to do it on the server. Particularly in markets that are less developed, the average revenue per month per user is quite low, and you can't afford to have expensive systems. Often an embedded device is sufficient for an insurer or fleet company.

We also support reverse geo-coding. Rather than reporting longitude or latitude, we can say you are now 200 meters east of a location. We have a customer doing this now.

Fleet owners or insurance companies are interested in finding stolen property, so the geofencing features are quite handy. Rather than having one or two circular fences, we can run thousands simultaneously or run complex polygrams. The device can calculate this inside the box, and you get an alarm when the box gets outside the area.

Does the device include an expensive chip to handle all that processing?

No, it's a small Cortex M3 processor. The history of this stuff is such that the original of this came out of Fastrax; I used to be its CEO and president. Navigil acquired the business from there. That application used spare CPUs capacity in the GPS receiver to calculate; we're used to managing with a very small memory budget. But now, the cost of a GPS device has gone so low that saving that dollar or two on the CPU is no longer significant. On other hand, the ARM7 architecture gives a lot more flexibility in terms of how much memory we need. And it doesn't tie us down to a single GPS provider. The intelligence is in the software.

Track-and-trace might be seen as a mature technology and market. Will innovation in this sector mostly be around lowering costs and improving efficiencies?

Both ways. Looking at insurance companies, our solution would reduce their cost of running the opportunity. To date, the UBI market has been cherry picking the most lucrative customers. When penetration gets larger, the pressure on cost becomes higher. On the flip side of the coin, if you want to increase the market, embedded intelligence, so you can do everything inside the box, gives you a low-cost method of determining if a driver is behaving well. Once developing countries get into the UBI market, they will carefully look at the cost of the box and the solutions being deployed to support this.

Susan Kuchinskasis a regular contributor to TU.

For more on fleets, see Industry insight: Fleet telematics.

For the latest on fleets, visit Telematics for Fleet Management USA 2012 on November 13-14 in Atlanta.

For all the latest telematics trends, check out Telematics Japan 2012 on October 9-11 in Tokyo, Telematics Munich 2012 on October 29-30, and Content and Apps for Automotive USA 2012 on December 4-5 in San Diego.

For exclusive telematics business analysis and insight, check out TU’s reports: In-Vehicle Smartphone Integration Report, Human Machine Interface Technologies and Smart Vehicle Technology: The Future of Insurance Telematics.

Services