Indoor Mapping is About to be Revolutionized
A few weeks ago, I posted a tweet about Google's "Project Tango":
This is the kind of mind-boggling, "insanely great" technical advance that Google is righly famous for.
And there are lots of other advances happening concurrently now with LIDAR and other technologies for rapidly building incredibly accurate 3D models of buildings - both inside and outside.
For just one example of this, take a look at Zeb1 / Zebedee, where one person using a handheld scanner can build a 3D model accurate to one centimeter in the time it takes to walk the hallways:
This is the kind of technology that makes a company dedicated to Indoor Mapping and Navigation salivate! We are committed to exploiting such technologies and approaches to make the most accurate Indoor Maps, more efficiently and effectively than any other Enteprise service provider.
But Data is Not Information - and Definitely not a Map
As excited as we are about this coming revolution in Indoor Mapping tools like Project Tango, it is critical to understand that a 3D model is not a map.
Let's look forward a few years from now (maybe less). Using Tango or a similar technology, Connexient will be able to send our map "go team" out to a client facility and build a 3D model - all the actual dimensional data along with complete photographic capture, dynamically married to GPS X/Y - and building Z - coordinates of a very large building in a matter of days.
That will be a fantastic tool for building a comprehensive 3D data set and representation of a building. But a user "walking" through such a model will be just as lost as the one walking through the real building. This is data capture for mapping, not the map.
In the same way, the satellite photographs that Google (and others) use also is not the map. It is very useful visual information that has many powerful user applications. But that is not the map, and definitely not a data set for turn-by-turn navigation.
It is only after we have captured comprehensive, accurate data that the work of mapmaking required for Indoor Navigation and Location-based Services actually starts. This involves challenges like adding meta-data that is relevant, eliminating extraneous data that is unimportant to the user, correlating data on POIs, places, people and so on to the map.
That requires - and for next 5 or 10 years at least will continue to require - a lot of hard, detail oriented, painstaking work by human beings. And keeping the data accurate requires good tools and techniques for both map-checking and making it easy for users to report errors and changes as they are encountered.
3D is Lousy for Maps
This may seem like a funny thing to say for a company that prides itself on our rich, interactive 2.5D Indoor Mapping and Navigation UX. But its true, and the distinction is crucial.
My brother has worked for 3DVia, a 3D modeling start-up and then acquired division of Dassault Systems for the last 10 years. Dassualt is among other things the equivalent of AutoCAD for building airplanes, factories, refineries, etc. So he knows a thing or two about 3D.
To summarize many discussions over the last couple of years - and his experience over the last 10 - tools for 3D have come a long way, but nearly far enough for anything other than very high end Enterprise and Entertainment types of applications. The problem is fundamentally not technical.
Once you make the jump to "3D", you are:
making a commitment to a level of detail in data, design & everything else that creates prohibitive cost;
You are adding resource load onto the CPU and RAM that is still beyond what is generally available in mobile devices or PCs;
generally adding complexity that is beyond most general purpose computers - it has been very hard to get 3D to work well except on dedicated specialized hardware and oftware.
To understand this, take a look at Panda3D for example:
The minute you go 3D, the user is conditioned - thanks to games & movies - to expect photorealism. No matter what technology you use to capture raw data, doing this well is very difficult and very expensive. No doubt the end result is super cool and compelling. But not an efficient way to go about mapping.
In a Map, Less is More
The key reason for this is that when it comes to maps, less is more. A map needs to reduce information to what is important and necessary for the user objective. In most cases, this is getting somewhere efficiently.
When you then combine map information with turn-by-turn directions in the carefully synchronized dance ofnavigation, the need to reduce information becomes even more acute. The challenge here is to provide the right information - and only the right information - at exactly the right time so that a user can make a navigation decision without distraction and information overload.
That's why Connexient selected the "3D-like" UX for our Indoor Maps. This approach - especially the "birds-eye view" that gives the user context of what comes next - adds real value to the map and navigation UX - but does that without the overhead of true 3D.
3D Will Become More Accessible, But . . .
There are two curves converging to an intersection sometime in the next decade that will make 3D more viable and accessible for non-gaming and filmaking applications.
Technologies like Zebedee and Tango that are dramatically increasing the efficiency of gathering the data / doing the 3D modeling of the "real world"; and
the continuing increase in processing power & resources on devices coupled with slow improvement in the 3D software platforms.
When these curves cross, there is no doubt user interest and demand for 3D will increase dramatically. But what this is not going to do is decrease by much the cost of producing 3D models that look good and feel real.
The key question for Enterprise Indoor Mapping and LBS is that of cost versus the user and business objective. What purpose does a 3D model serve?
For Indoor Navigation 360 Degree Photo Models Will Dominate
There are already lots of good examples out there of how this kind of complete virtual photographic model of a facility has other valuable applications in operations, security ("situational awareness") and so on.
And when it comes to the Indoor Map, the concept of Google Street View with 360 degree navigable interior photos is obvious, and one that Connexient is already pursuing. That type of immersive representation of the real world - when and only exactly when it is needed by the user - is very helpful.
Combine this with the ability capture, geo-locate and reference such photos cost-effectively, and it is a very powerful addition to our Indoor Mapping solution.
But Is Mostly Irrelevant to Indoor Navigation
It all comes back to that simple truth: when it comes to Indoor Maps and Navigation, less is more. The last thing somebody walking down a hallway needs to be looking at is the picture of the hallway they are walking down. To pre-plan a route or confirm where you have arrived at the right place? Sure. But not while in transit.
What you need while in transit is a simple, intuitive set of choreographed visual and audio cues of when to turn where, and what important information that can be easily observed in the world around you can be correlated to that.
Let's again step forward a few years and think about a world, perhaps, where millions of users stroll around with Google Glasses on.
It is easy to envision then that the visual representation of the map goes away entirely. The user will not even view a map, much less a 3D model. Rather, the will have the map and navigation information and cues projected onto the real world they are observing and navigating.
While the Map Remains Critical
The other simple truth - and Connexient's mantra - is that the Map Really Matters. The different ways of visualizing a map are also important. But at the end of the day, it is the accuracy of the underlying map data and metadata - and how it is correlated - that make it useful.
No matter how compelling, beautiful and elegant the UX, if the map does not get the user infallibly to where he or she wants to go in an intuitive way, it will not be used.
So we remain eager and enthusiastic early adopters in every technology that will make that job faster, more accurate and less expensive - while keeping a perspective and focus on what will actually be useful, not just cool!