Check Engine!
The car is designed by them for us. It needn't be and shouldn't be.06-Jul-2002

CNN reports that Automakers are often unwilling to provide dealers with the information necessary to diagnose problems in cars. http://www.cnn.com/2002/TECH/ptech/06/24/diagnosing.cars.ap/index.html

This is a problem I'm painfully familiar with. The week before my "Check Engine" light came on. I had worked with the AutoPC group at Microsoft and am aware of the ODB-II interface. ODB means Onboard Diagnostic Bus. It is mandated by the US EPA (Environmental Protection Administration) so that cars can be monitored for adherence to pollution regulations.

Since it must be present anyway and is a common standard, the bus has been used for many other purposes within the car including controlling the windows. It is the way that the engine reports information about what is happening in the car including a (limited) history of problems including why the "Check Engine" light came on.

When the Check Engine Light came on (while driving to and from the Massachusetts Broadband conference -- coincidence?) I continued to drive and then called for an appointment to find out what the problem was. But by time I brought the car to the dealer the light had gone off and it didn't seem to be worth leaving the car there. But it came on again the last week and I had no choice as I was going to be driving a few hundred miles that weekend. I had to bring the car in early in the morning before my son's elementary school graduation and hope they could diagnose the problem and repair it so I could fly to the INet2002 conference later the same day.

I was annoyed at what I consider the simple refusal to share information with me. It's like replacing a temperature gauge with what is called an "idiot light". It turns red when something is wrong whereas with a gauge I can drive my car carefully and keep the temperature within a safe range until I can get it repaired.

While the separation between content and transport in connectivity is a fairly simple boundary, the boundary between what I can and cannot do in a car is less clear. In fact, in investigating the ODB-II situation I did found places where I can buy my own diagnostic aids including one that displays the information on a PDA and allows me to analyze it on my PC. Car hacking used to be about mechanical alterations and is now often more about software than wrenches. I suspect that these independent ODB-II vendors are much less expensive and possibly more capable than the versions used by approved dealers since they have to deal with a more active marketplace.

The automakers have a long history of trying to be the provider of all that is in the car. For a while, however, they didn't do a very good job and the result was a thriving aftermarket in consumer accessories including improved radios and entertainment systems. The aftermarket still exists, though is smaller now that the built-in systems have improved and the CB radio market is not what it used to be.

Building devices into the car makes sense as long as they are unlikely to change and it is necessary to tightly couple the device to the car. There was a period when cellular phones were built into cars but by the time they had become pricey accessories, hand held phones were becoming common and it just didn't make sense to build one into the car when you needed it with you all the time and it can fit in your pocket.

As an aside, Bluetooth (http://www.bluetooth.org) harks back to those days and is designed around the assumption that you'll have only one cell phone that is the gateway for all your devices. I already have a separate "phone" on a PCMCIA card in my laptop. I consider it an expensive experiment since the per-radio pricing is reminiscent of the days when phones were priced according to the number of copper wires that came to your house. OK, that is still the policy but I'd prefer it to be a reminiscence.

The concepts of connectivity that apply to telecommunications apply just as well to cars. There is no longer a need to build every device into the car. The display surface for a device, such as a radio, doesn't need to be located near the actual tuner. We already see this with CD changers that are located in any available space in the car.

Car navigation systems are available stand-alone, as add-ins and are now often built in and it will be interesting to see how this marketplace evolves.

Imagine that we had a general purpose (IP) network in the car. As new options become available I can take advantage of them. Devices can be self-sufficient or take advantage of sharable components such as wireless connectivity (cellular, 802.11 or whatever becomes available), location information, car instrumentation, display surfaces and control interfaces.

We don't really have to imagine it. I can just take a small computer and my Verizon Express interface and put it into a car with an 802.11 network. I can then connect my iPAQ to the network. I could then share navigation information with other cars and have a far better dynamic map of the traffic conditions than I can from the periodic radio reports and I can integrate the information directly onto my navigation screen. With Internet radio I can listen to radio stations around the world or my favorite information sources. Instead of relying on tightly controlled and defined services such as On-Star I can subscribe to any assistance services available.

I can do some of this but haven't. And there aren't the wide variety of services that I would want to use. Internet Radio would be a much better approach than Satellite radio since it would give everyone an opportunity to broadcast, not just those who are in hock for a satellite. Too bad that Internet Radio has been deemed too disruptive to be tolerated. For now, I use my built-in or iPAQ navigation, listen to NPR (and sometimes other radio stations) and simply try to avoid driving during rush hour and work at home.

What will it take for the car marketplace to mature? The market will evolve -- the idea of driving changes from the edges is too compelling and we do have the tools to be able to do it. The question is what path will it take and how long.

There are some enabling technologies including encrypted IPV6 that are needed to make pervasive connectivity simple. But the real issue is the lack of building blocks that allow us software-types to redefine the car. At this point I expect that we'll see PDAs evolve into more general purpose computing platforms.

PDAs started out as special purpose devices since they pushed the edge of what was feasible and discarded any capabilities not necessary for their initial purposes. A symbiotic relationship with the PC was part of this. Palm computing was originally a software company that had to build the hardware because none of the hardware manufacturers were ready to build the necessary devices.

The car is a relatively contained environment and thus a microcosm of the changes that are occurring in the world at large. It took nearly twenty five years for the early packet networks (such as the ARPANet) to evolve into the Internet we know today with the World Wide Web as a common platform. In the next stage we can build upon the edge-driven concepts with connectivity giving us an ability to separate control from topology. But as we've seen with the PDA, sometimes vertical integration is a necessary stage. In fact, the early days of telecommunications had the same whole-system design characteristic. But it is only a stage.

The Internet tamed the telecommunications infrastructure and gave us connectivity so we software types could build upon it. I look forward to the ability to have similar control over my mobile environment.

In the meantime, I may indeed get desperate enough to buy an OBD analyzer though I realize that it will only give me a hint of what is possible. After all, what good will it do me to know that the frobnitzer level is 17%? I apologize if frobnitzer is a word in any language and, worse, if it is an obscenity -- I don't have the staff to check for such possibilities but, then, the GM No Go (Nova in Spanish) shows the difficulty of avoiding such problems. All the more reason to share the ability to add value and not assume that one omniscient provider can or should do it all.

As long as car companies see the aftermarket as their only source of profit they have little incentive to be a creative force. The larger opportunity will be to redefine the car. Perhaps that will happen. The Swatch-car as a harbinger (you effectively build your own car) and electric power providing some freedom from mechanical constraints.