A few weeks back we had a flooding event here in Carbon County, Utah. A hotline was established and folks started calling in for help and to report damage. My GIS department was tasked with the usual chore of mapping the locations of all the calls to help coordinate the response.
Folks called in with a name, phone number, address and a description of the damage. While most of the time our mapping duties went smoothly, we had the usual difficulties we all face in situations where an address is the primary location information. First, we had Haycock Lane, which is also 3500 West. No problem if you happen to know about it—look it up by the house number and then consult the roads layer. Then there were the more tricky situations where they either reported the wrong address, or it was mistyped when it was entered. But being spatial specialists and also having the added benefit of “local knowledge,” we could look up the name in the parcel database and figure it out from there. However, it was a little more difficult when the address was wrong, and it was a renter who called in. Furthermore, it was tough to tell how many calls were separate incidents when both the owner and the renter called to make a report, especially if the reports didn’t quite match. (The same damage may represent different value to different people … “beauty in the eye of the beholder” and all that.)
Then there were descriptions of the location and potential damage that at first seemed simple but turned out later to be more difficult to locate. For example: golf course, ”X” dollars of damage, then tie that to the address point at the clubhouse, check; power company, address point at the plant up the canyon, check; railroad, train station in town, check-check. But then the Federal Emergency Management Agency (FEMA) shows up and ruins the party. They want specific locations of all damage. The golf course clubhouse is not damaged nor is the power plant itself. The damage is located at various places on specific parcels but still a significant distance from the address point. And you guessed it, the train station is high and dry. Actually, the railroad’s reported damage is a section of track just behind Joe Smith’s house at the end of Mythic Lane, you know, on the west side of the field, kind of kitty corner to . . . okay, EVERYONE knows the field with the old red tractor behind there. The tracks were washed out by the old milking shed—about a hundred yards to the south.
Sometimes, an address just doesn’t cut it—at least not in this situation. But to understand some of the complexities, we need to ask the big question: What is an address? While I might not be able to answer that for everyone in every situation, I can give a general answer. An address is simply an abstraction of a location. However, to be useful it can't be entirely abstract, so we tie it to our system of roads. Typically we number either the structures along a given road or the distance down a road in a given direction from the origin. I don’t know about your county, but in our county, we do not address along “rail” roads. Further, we don’t address anything by county ordinance but “addressable structures.” That leaves the train station washout incorrect on two counts. A traditional address, were it to be issued, would not work here because it is not on a road. What we really need is a location. Why? Because we wanted to map it and computers require coordinates.
Now, I know some of you are starting to protest. After all, you can put your address into Google Maps and it will give you a location, you say. But that's only because a service called “geocoding” is operating its black magic in the background to make the translation. Geocoding is often incorrect in the rural areas where roads don't lie neatly on a grid. But beyond that, geocoding is limited to the areas that have been addressed, meaning areas near an existing ranged road. Disasters are notorious for not consulting the map to ensure they occur near roads. They will drown your crops in the middle of the farm. They will burn the mountain side. They will break up your vessel in the middle of a body of water. Addresses are no match for disasters.
So, if addresses are such a problem, why have we not done something about it already? For one thing, we've always counted on addresses (both sites and mailing) to differentiate between locations. You know, good old “always been done this way.” Also, everyone has a stake in addressing, meaning we've all memorized our addresses, and we have all the documents and services that define us and our things tied to those addresses. There's a tremendous mass of public knowledge tied up in addressing, and (as all masses do), it has an equally tremendous inertial resistance to change. Parting with this tradition requires tremendous energy. If there are a thousand addresses to change, there are a thousand people who each have several reasons why it shouldn't be changed. It would seem then that we should just give up on any idea of making a change.
However, that's only part of the issue. As is often the case, we insist on relying on old technology that becomes increasingly less equal to the task at hand. In the past when there was an ambiguity in an address here and there, what really was the consequence? So you got your neighbor’s package? In all reality, you most likely have to admit that you've needed a reason/nudge to go talk to them anyway. No big deal. But as we tie more and more to an address, we have greater problems with the ones that are in error. Now it isn’t just your package that fails to arrive on time. Now you have trouble getting homeowner’s insurance because the insurance company’s geocoding processes don’t find your house. Or you have trouble getting your home loan for the same reason. Or you can’t vote for your political leaders when you show up at the polling location because the geocoding service didn’t identify you as living in a particular precinct.
I call this address abuse; because when the person or system assigning the address completed the task of geocoding, they or it had no idea the assignment of that address would be making those types of determinations. Addresses often lack the spatial resolution to reliably find your parcel. They were never meant to make accurate assessments about which voting district you live in or whether or not you qualify for homeowner's insurance. And the abuse is only going to spread with time.
Now maybe you are still of the opinion that I haven't made a particularly compelling argument. Perhaps so, when you consider each example in a vacuum. But remember, each argument adds to a combined inertia of its own in opposition to the argument to let addressing stand as it always has. We live on a continuum somewhere between the Neanderthal looking for his next cave and a Star Trek Enterprise crewmember getting beamed down to the surface of a never-before-seen planet. While we haven't made it to the Star Trek end yet, with the looming possibility of instant package delivery via drone, the misdelivered package issue may become a much more frequent problem, not a chance-to-check-up-on-your-neighbor problem. Moving forward, automated services—virtual and physical—needing accurate location data will become the norm. The explosion of location-based information combined with the coming Internet Of Things (IoT) boggles the mind. It also makes current addressing solutions look about as equal to the task of location as the Neanderthal’s club seems capable of going head-to-head against a phaser. But we will likely continue to try to differentiate ourselves with addresses until we just can't do it any longer. That is the cycle of things.
The madness comes in the myriad geocoding services we'll employ to translate the abstract address we insist on into locations that can be used. Think about the fact that every local government typically has the authority to set its own addressing. This often means a dozen city address grids for each of the 3,144 counties in the United States. Each of the tens of thousands of municipalities also has the liberty to set its own address grid origins and create its own addressing systems. Now add to that the number of addressing systems each sensor manufacturer will come up with in the IoT, and we'll need a supercomputer just to correlate the spatial data. It'll be a death of a thousand cuts, and we will have inflicted each one on ourselves.
The answer, of course, is simple. After all, if an answer isn't simpler than the problem, it probably isn't really the answer: Make the leap to locations and get rid of addresses entirely. The whole earth was “addressed” centuries ago. It's been standardized for decades. It’s latitude and longitude. The reason we don’t all use it is that, in the past, it took special training and equipment to locate yourself, but not any longer. However, latitude and longitude addressing is not very user-friendly in its native form. The numbers are longer than we can easily memorize and there are the three different formats, which you cannot easily convert in your head. But if you use the decimal-degree format for everything, you can do a couple of things to quickly turn it into a user-friendly grid while keeping it compatible with a computer-friendly location. First, the degrees repeat themselves only every 69 miles or so. In most cases we can ignore them since city, ZIP Code or nearby landmark will locate a person, and the computer can remember the long numbers. Then, for house “locations,” we need (in most places) only the next four digits or sometimes five in one direction. If someone uses five digits in both directions, they are likely locating some form of infrastructure, like a light pole or manhole. If someone uses more than five digits for both the latitude and longitude, they're most likely locating things that are survey-accurate.
Getting back to my flood example, depending on the significant digits used, there's already an "address" we can all understand for each individual crosstie of the washed out railroad, if need be. As for mistyping the directional, there are only two that are valid in this hemisphere—north and west, in that order (mathematicians, see me after class). Addresses become standardized, unambiguous and unified.
Is this system a panacea? Definitely not. But if this is the weapon we choose to unify the communication of location information, we will find ourselves much closer to the phaser than the club.