Navigation systems: why they're lame, what they need

OK, I'll do the dirty work: In-car navigation systems are a flop.

There, I've said it. I've said what nobody else around CNET's tech-crazy halls wants to admit. In-car navigation is the Sony Watchman of the car business: a clever video gadget that doesn't solve a real need. And the paucity of buyers is proof.

Think about it. Most of us drive to the same familiar places each week. We don't need directions. And when you do go somewhere new and unfamiliar, with today's navigation systems, you have to pull over and enter the address or race to tap it all in during a red light, thanks to lockout technologies that insult the user. (For example, I just tested a Jaguar XK-R that reached a new pinnacle of annoyance. You have to press a button just to get to the nag screen, then you have to press a button to acknowledge.) And when you do take the time to enter a destination, the system rewards you with a clumsy route that no local resident would ever take.

Recently, real-time traffic data has been arriving on the car navigation scene, overlaid on your map as you drive. Sounds promising, right? Actually, not really. The data set is way too coarse to be useful, and there just aren't that many sensors in the roads, if any at all. Besides, that little pea-brain nav computer can't begin to figure out a sneaky way around a traffic problem; it doesn't really know your town.

Plus, traffic data is like a typical weather report: You can't do much with the information, but you still want to know. You'll still go through your day, the only difference being that when you're soaked, you expect it. It's OK to enjoy that feeble form of knowledge -- just don't confuse it with empowerment.

And don't even get me started on those lame points-of-interest (POI) overlays in car nav systems. Surely, they can offer something more useful than icons showing all the churches and parks along my route -- handy if I decide to pull over and go to the park restroom, then need the shortest route to find a confessional, but otherwise, no.

I know this all sounds bleak, but I haven't given up on navigation systems entirely. I've been watching several advents that can fundamentally change the nature of in-car nav and save it from being a huge in-car flop. Here they are:

Learning systems: Pioneer's upcoming AVIC-Z1 system could be an early example of nav systems that "learn" to be your copilot. By monitoring the streets you drive for a while, units such as this can delight drivers by suggesting routes that incorporate the roads locals use with the greater comprehension that only a more advanced computer can muster. Finally, you can look forward to a nav system that teaches you a thing or two about driving in your town, rather than just hoping it can keep up with you. That said, I will admit this is a subtle advent, one that may be lost on prospective buyers until they experience it. To succeed, this technology will need massive word of mouth from early adopters.

Peer-to-peer traffic data: Check out the white paper on the TrafficView (PDF) system developed at the University of Maryland, and you'll see a future with dramatically upscaled traffic-data resolution; every car on the road becomes a sensor and a reporter. Better yet, the peer-to-peer nature of it avoids the need for a massive traffic-data network build out that nobody wants to fund. The biggest hurdle I fear here is interoperability. Carmakers are loath to offer anything that works with their competitors' vehicles. So this vision may have to be carried out with clunky, add-on devices. That's a shame.

Dynamic POI overlays: Churchill Navigation is showing a very slick navigation system that combines a photo-realistic display with super-rich overlays. The company even custom-wrote one that shows all the abandoned missile silos in the country for a client who is a buff. Combine that niche focus with the possibility of dynamic data, such as new local restaurant reviews, daily theater listings, business licenses, and -- yes -- advertising from merchants you patronise, and you can imagine a POI system with a living, breathing set of points that reflects your life and interests. Unfortunately, Churchill probably won't be the company to do it. It admits it wants to sell a very expensive system to rich guys, not focus its business on dynamic POI subscriptions.

Overall, today's in-car navigation computers suffer from a fatal flaw: They pale in comparison with even the oldest PC in your house. Smaller, dumber, and more limited -- not to mention boasting a worse display -- the in-car nav system needs to get better, and the price has to come down 75 percent (that's another column). However, with these three technologies, I think the relaunch of in-car navigation could be a hit.

Do you think today's in-car nav systems are a joke? What would you do to improve them? Talk back below!