Internet ITS Consortium IP Addresses for Each Car-mounted Device

Naoki Tokitsu
Chief Secretary


Internet ITS Consortium of Japan is an industry group actively seeking technology development and commercialization of Internet ITS, which integrates Intelligent Transport Systems and the Internet. Naoki Tokitsu, Chief Secretary to this consortium, was involved early in the development of car navigation systems and Japanese Electronic Toll Collection (ETS) system. He is also a believer of possibilities that IPv6 brings. IPv6style asked Tokitsu the relationship between IPv6 and automobiles.


Unlimited contents to Every Car

IPv6style:Can you define ITS?

Tokitsu:A lot of people ask me what ITS is. In Japan, VICS (Vehicle Information and Communication System) and ETC are two good examples of ITS. VICS sends information through communication infrastructure to show routes and traffic information. ETC enables toll payment through communication with toll gates. I would define that ITS is anything that is based on communication between cars and other entities.

Some people are conducting research on automatic braking system in which cars would have radars to detect automobiles or obstacles ahead, and call such system as ITS. I don't think that is appropriate. Such systems don't have any communications with obstacles ahead. To be called ITS, there should at least be a mechanism for automobiles ahead would send back signals for the system to receive. If the car stops on its own, that is just an extension of existing system, and not ITS. It should be a little clearer if you make distinction between the two on that point.

IPv6style: Then what is the difference between ITS and Internet ITS?

Tokitsu: ETC is only used for toll collection, and VICS only shows you traffic information. Internet ITS aims to go beyond such "only" world, fully utilizing the power of Internet by combining ITS with the Internet. That enables myriads of things, such as listening to downloaded music and uploading specific automobile information for diagnostics. Internet ITS Consortium is seeking how we can build adequate business model. We would like to share how we can build business out of connecting automobiles to the Internet.

IPv6style:There are already those contents services that enables download of map information through the Internet.

Tokitsu:With Internet ITS, we are trying to make the unlimited Internet contents available to automobiles. The more contents available, the better. Some automobile manufacturers offer services that allows download of information to car navigation systems. But they have not become so popular, because few contents are available. They may charge only 600 yen a month, but they can't find many users, because people don't think these services offer contents that match the charge.

But some people pay 20,000 yen a month for mobile phone Internet service. People find 20,000 yen monthly value in mobile phone service but 600 yen is too expensive as automobile service. We considered why that is so and we found that current automobile services don't allow users to utilize the unlimited contents available on the Internet.

Some people object to this idea. They say that allowing drivers to use the unlimited contents on the Internet is dangerous, that it would lead to accidents. Then why not produce the kind of cars that allows drivers to watch TV as much as they want and still avoid colliding into the car ahead. Cars should communicate with each other and control speed if they get too close. Of course, this is not possible without high diffusion rate of such systems. That's why some say it's not possible. But we can imagine that any automobiles will have similar systems 10 or 20 years from now. It can't become popular overnight. We are trying to start from what we can. Right now, we are developing systems to connect cars and the Internet. Once we get connected to the Internet, we can use contents worldwide.

An automobile manufacturer offer contents services because they want current owners to buy its product again. But they only allow the customers to connect to its information only. Users are not willing to pay because they offer closed connection. I think car manufacturers to compete with contents or applications on the terminals, rather than connection.


Move to ITS Begins with Car Navigation Systems

IPv6style:How are we going to see move to Internet ITS? Some current car navigation systems can receive e-mails. Are we going to see more general-purpose terminals with more contents available?

Tokitsu:Car navigation systems can only do car navigation. We pay 200,000 yen to a device only capable of car navigation. With the same amount of money, we can buy a portable PC. Portable PCs have large displays, and have all communication functions including wireless LAN. Car navigation systems don't have those and still cost 200,000 yen. Ideally, PCs should be on automobiles, as far as hardware is concerned. Car navigation is essentially a software function, so such software can be run on PCs. Ideally, car navigation systems should transform to PCs. That is one of our targets.

IPv6style:You are saying that current car navigation systems should move to PCs, including OS. It must not be difficult technically, but how far is it from getting real?

Tokitsu:Before we reach this goal, we have to take care of current car navigation systems market with annual shipment of 3 million devices. This is a several tens of billion yen market, and it can't be changed overnight. General-purpose OS can be used for these systems and car navigation can be implemented as software. Everyone thinks these systems should transform this way in the future. But now they are competing with proprietary systems and they sell. We can't overturn this market at once, so we need to see what we can at every car model change that comes every two years. The 2005 model is in production, and manufacturers have already fixed on car navigation systems for next 2007 year model. So we need to target 2009 year model.

So it takes at least four years to get a device with new functions on a car. We are counting on business vehicle market, such as trucks, buses and taxis, for earlier changes. Current business vehicles use passenger car products for car navigation systems. These systems can guide routes, but they can't serve additional functions such as delivery order instruction. We are looking at getting general-purpose terminals connected to the Internet in this business vehicle market, and establish business, for later implementation on passenger cars. We would start in 2009 for passenger cars and car navigation systems might be changed completely in 3 generations or 6 years. In other words, we may see general-purpose vehicles that can do anything to flourish in 2015.

IPv6style:What type of services do you think will catch on first?

Tokitsu:There are some services already getting popular, such as remote maintenance, telling when to change engine oil, for example. They are combining this service with insurance, offering discount on insurance fee if the car can be managed via remote maintenance. After-market service will be an earliest example of successful business.

The most difficult will be contents download business. It would be fun to be able to download movies and music, but forming business around it would not be easy. Promising business is where it is closer to automobiles.

IPv6style:With Internet ITS, network connectivity should be a very difficult issue. Cars need to switch among various network services such as cellular phones, DSRC (Dedicated Short Range Communication), and wireless LAN. I imagine there will be a difficult technical challenge to manage those connections.

Tokitsu:That's where IPv6 can help. Currently, if you get connected with cellular phone, you get connected to it and nothing else, and there is no way to switch automatically to DSRC, for example. If we could use Mobile IPv6, you can build constant connectivity. That is our ultimate goal.

IPv6 infrastructure is not available yet. But Internet ITS Consortium develops our systems on IPv6, while using IPv4 for field trials. We can always move to IPv6 if IPv6 infrastructure becomes available. But for such ideal infrastructure, there is a need for a connection provider that takes care of connectivity aggregation. Such provider may successfully transform to an integrated automotive portal, attracting access from 70 million cars. We can't allow someone to monopolize on it. Such function must be managed on non-profit basis jointly by those concerned.

IPv6style:What other issues do you see in network connectivity?

Tokitsu:It's the service availability. You can use cellular phone and DSRC, but we also expect a lot on wireless LAN. Wireless LAN could provide for 50Mbps no cost connectivity. Wireless LAN components are so cheap that you can find them almost in every portable PC. So necessary components are there on PCs at least, and infrastructure allows free connectivity. The issue is who would want to offer such free connectivity?

This issue might be overcome soon. We conducted field trials in March and June 2004. At both times, we placed wireless LAN components at vending machines. Vending machines have power supply. You can get this electricity and place antennas on these vending machines, to set up hot spots instantly. Connecting each one of them to terrestrial Internet costs a lot. Instead, we can make vending machines communicate with one another, to create a "hot area", offering free internet connectivity. We built such hot area in Nagoya, Japan. At field trials, we experienced getting stable connectivity once we got into hot area, while we had to put up with intermittent communication using cellular phones. We even did cost simulation. It takes only 50 million yen if you were to place 100 access points in 1km radius. For 50 million yen, such services can allow Internet connectivity to retailers, cars, bicycles, walkers and offer free phone connectivity through VoIP.

IPv6style:You said Internet ITS development is based on IPv6. Would one address for an automobile be enough?

Tokitsu:No. IPv6 has two advantages. One is connectivity, and the other is ID. For example, engine ECU (Electronic Control Unit) could be given IP addresses for monitoring. You could even change parameter settings. Meters for automobiles are controlled by self-diagnostic function. Automotive multi-functional meters are controlled through diagnostic systems. Each of car-mounted devices should be given IP address for abnormality tracking. For parts recalls, you can see which cars have already changed necessary components. We could also tell users on unrefreshed parts.

In automotive industry, we have a term VRM (Vehicle Relationship Management), instead of CRM (Customer Relationship Management). You take care of cars well after sales, and user would buy at the same dealer. For such services, you had better be able to collect information from each car part. Therefore, 1 IP address for a car is not enough. If you have 1,000 components on the car, you had better assign 1,000 IP addresses to them. IPv6 as ID offers a great possibility for future cars as well as Internet ITS.