Exploring Local
Mike Dobson of TeleMapics on Local Search and All Things Geospatial

Screech, Bang, Boom – I love my navigation unit

April 22nd, 2010 by MDob

Today, Google announced that its navigation service in the UK was going live. Imagine the number of accidents that will cause? Of course, since Google announced driving directions were available in an additional 115 countries earlier this month, I have an obligation to advise all of you to avoid buying new cars during the next 12 to 24 months. Junkers should be ok and very practical, now that everyone will be going the wrong direction.

The other day Duane Marble sent me a quote that had appeared on the Directions Magazine website about a problem that the Department of Transportation in Foxboro, MA was having with drivers paying attention to their navigation systems and not paying attention to the layout of the roadway ahead of their cars. See this for the full article from the Sun Chronicle.

In this case, the road described is a dangerous one for residents, since drivers seem not to notice a stop sign at the bottom of a local T-intersection. Instead, for some unknown reason, drivers continue straight through the intersection, driving over a stone wall and colliding with cars parked in the driveway of the house across from the stop sign

Previously Foxboro had problems with Google Maps, as the service had somehow mislocated Gillette Stadium (perhaps that explains the Patriots’ lousy season), although they managed to show the location of Schaefer Stadium, whose name had been changed to Foxboro Stadium in 1982. Unfortunately, Foxboro Stadium was demolished in 2002, when it was replaced by Gilllette Stadium. However, if you search for Foxboro Stadium, you can find its name on Google Maps, next to Gillette Stadium. Of course, if you search for Schaefer Stadium, you can find it at the same location as Gillette Stadium. Perhaps, of more relevance, Google had located Gillette Stadium in the middle of a small subdivision that, on game days, was crushed with traffic headed to the wrong destination.

Getting back to the more recent problem, Foxboro Superintendent of Highways, Robert Swanson said, “We can make the street one-way, but GPS and online/Web driving technologies will still consider it a two-way.” The Superintendent went even further, suggesting that the time may have come to ask the Massachusetts legislature to “determine if some form of law or statewide clearinghouse can be formulated” to require GPS and online traffic services to update their maps soon after towns submit traffic changes. Swanson added, “This appears to be a systematic problem that is not being addressed at the state or federal level.”

I called Mr. Swanson to discuss his idea, but was told he was dealing with his IT group about computer problems – I suppose they had tried to install the McAfee never-ending reboot program (the endless reboots are apparently a new and highly desirable feature among workers looking for an unexpected day off).

I know a lot of you probably find Swanson’s call to action a bad idea. I, on the other hand, think it has merit.

Take Google Maps, as a good example. If you or I think we have found an error, we can go online and submit a correction. If you are an authoritative source, say a municipal highway department, or a State Department of Transportation (remember the Providence, Rhode Island debacle) you can…well, you can get in line just like everyone else and submit a correction to Google. Then, you can submit one to Bing, MapQuest, Yahoo, NAVTEQ, TeleAtlas and one to every other commercial vendor providing navigation services or mapping information used for navigation. Of course, if you go to Garmin with this type of request, they will, as they recently demonstrated, tell you that they are not responsible for the quality of the map data they use and that you should contact the vendor.

It seems curious that the best way to solve this problem, at least in terms of personal satisfaction, is go to OSM and fix it yourself. Hmmm. Good solution, but that Creative Commons license will likely keep it from ever being exposed to enough users to solve the problem we are discussing here.

It seems to me that if you are an authoritative, government source responsible for highways, addresses, or other geographical data and a map error is causing a problem of public safety, there should be a method of remedying those problems faster than the four months our mapping and navigation colleagues have taken to resolve the problem in Providence. As a matter of fact, the expansion in the number of these types of problems is leading me to suspect that local mapping companies are going to spring up to create quality “local mapping” and make the “big boys” look even more inept.

Oops – Just off the press

I have to rescind my comment about the “rapid four-month” solution to the Providence problem. See this image taken this morning of the NAVTEQ website, showing that traffic is slowing to 33mph on the section of I-195 that has been closed since November and is being demolished. Oh please, make this stop – fix this error. You can do it NAVTEQ!

Wow, traffic on the demolished road segment is slowing,  while the green band is doing 44 mph.  How about that?

Perhaps Providence is a good example of the types of errors that concern Mr. Swanson. It is likely that some of the companies providing navigation services that incorporate NAVTEQ data are supplying bad routes through Providence. In fact, if navigation systems users are driving from east to west on I-195, they will think that they need to merge right to connect with I-95 north or south, when they actually need to merge left to connect with I-95 south and keep right for I-95 north, but not on the road segment described by the NAVTEQ data. Since NAVTEQ clearly knows about this error, how can it take them so long to correct it on their own website? Oh, something about how long it take them to compile their data into a runtime version? Hmmm. I thought we had left the Middle Ages centuries ago.

This blog was just to soften you up. Next time I will direct my comments to establishing a road clearing house and why it might be an idea whose time has come. Perhaps, Mr. Swanson will have called back by then, if so, I will let you know what he had to say. In addition, I will spend some time describing my thoughts about local mapping, a market segment that may have a “second-life” after it was killed by online mapping and routing.

Click for our contact Information

Bookmark and Share

Posted in Authority and mapping, Google maps, Navteq, Personal Navigation, routing and navigation, TeleAtlas

4 Responses

  1. Duane Marble

    Just being an “authoritative, governmental source” does not mean that you have correct information on what is out there. The official map for our little town of Florence, OR comes from the Oregon DOT. (see egov.oregon.gov/ODOT/TD/TDATA/gis/docs/citymaps/Florence.pdf) This is dated 2008 and omits some street names and shows the police department at a location that they have not occupied for years. There is, of course, no metadata that provides any information on where ODOT got its information.

    The situation becomes even worse when these inaccurate base maps are used in conjunction with whatever mashup spatial data comes to hand. Think, for example of taking all the spatio-temporal traffic accident data from our 50 states (assuming implicitly that all accidents are reported according to rigid, uniform standards that exclude any local variations in reporting) and putting it on one of these inaccurate maps. Now, lets use this for some policy decisions . . .

    Thanks, Duane:

    While I am sure that you are right about the inaccuracy of some sources that we would think to be “authoritative”, I am sure that Rhode Island Dot could have provided the data on the I-195 realignment to any interested party six months ago. While DOTs may not be authoritiative in respect to every street (and I would not expect them to be, since they do not have authority in most states for maintaining every street), they certainly are the ones who pay the fees and hire the contractors to reaalign intersections. At the local level, it is the county/city/town highway authorities that have responsibility for changing street directionality, etc. We could think about providing a mechanism for them to distibtute the change material regarding the jurisdiction for which they have authority, or we could just continue to wait for the mapping companies to discover these things for themselves. It seems that something needs to be done to improve the kinds of problems that we are seeing.

    Mike

  2. Duane Marble

    Mike,

    I am in full agreement with you that something needs to be done here. A unified “change central” could eliminate the bits and pieces approach that has led to our current problems. However, the relevant local, county and state governments also face massive difficulties in keeping their street data current. Here in Oregon, we have just finished the first decade of a “tax holiday” put in place in the late 1990s. It has resulted in increasingly unpleasant budget problems. In theory, the City of Florence should keep a sharp eye on all its local street information and promptly inform the county and state of any changes or identified problems. In reality, there are no city staff available to do this since they are overburdened with other things that are more critical to the day to day operation of the city. I suspect that ODOT itself has problems of the same sort.

    The point of the above is that even with a “road and street data central”, the data acquisition process for the needed change information is most likely broken as far as the local government side is concerned. Quite likely, this will vary significantly with city size, status of budgets, etc., from place to place be we are unlikely to see much improvement. The United States has, I gather from reliable sources, well over 40 million links in its road and street network. We would like this kept current with relevant link attributes such as name, speed limit, number of lanes, stop light and stop sign locations, turn impedances at all nodes, etc. Somehow I do not feel that user generated data is going to help us very much in obtaining it. Is there a viable solution? I would hope so, but I can’t think of one!

    Duane

    Thanks, Duane:

    Your points are good ones, but I still think the concept is one worth thinking about.

    Mike

  3. Matt McGranaghan

    Computers, databases, and wireless digital networks
    may or may not speed the cartographic process but do not
    change the aphorism “When the map and the terrain disagree,
    trust the terrain.”

    Sales pitches and optimism aside, maps and their derivatives
    are abstractions — best treated as provisional
    statements — handy when they work, requiring continued
    re-evaluation and considerable effort to ensure that they might,
    and willingness to recognize it when they don’t.

    There is indeed a lot of work to be done.

    Thanks, Matt:

    So, if that word processor you purchased doesnt’ work, you should be “OK” with that, just as you would be if your PND provided a route that was illegal, unsafe or one that did not connect with the intended destination?

    I think this is a ‘fitness for use” issue and one that has specific implications for public safety. If there were a way to mitigate this problem, should we look for a solution or ignore it? Your comment that the use of maps is one “…requiring continued
    re-evaluation and considerable effort to ensure that they might,
    and willingness to recognize it when they don’t” might work fine when you are hiking, but you may not have time or the ability to react to the error, if you are driving a vehicle. However, you may be suggesting that maps cannot be used in certain applications like navigation because digital map databases are abstractions of abstractions. Hmmm. 😉

    Mike

  4. Jim Donahue

    Its amazing that both Navteq and TeleAtlas seem to ignore their
    own available gps traces from gps users in the US when both
    companies take advantage of that data where they do not have
    mapping vehicles on the ground. TeleAtlas apparently mapped
    parts of Russia using only gps traces from TomTom customers
    but they like Navteq have not completed the 195/95 interchange
    either.

    Certainly their software must pick up on user anomalies and
    decide when its the norm and that their current data is wrong.

    Hey why do we have 500,000 passive gps traces showing cars flying
    over water in Providence? Gee lets look into that.

    Thanks, Jim. You raise an interesting question.

    Mike