Strategic Teaming for IPv6 Applications
by Chris Harz

A great deal of attention has been given to the infrastructure and details of communications via the New Internet (which is all necessary and good), but relatively little has been said about the related industries that will be affected by v6 for upcoming applications. That's too bad, because much of the motivation for the growth of v6 in America will come from the symbiosis between the technical and applications communities. To draw an analogy, it would be useless to develop the world's greatest videogaming platform and then not have any games ready to play on it when it came out. Similarly, the explosive growth of the automotive industry at the beginning of the 20th Century was not just due to the heroic efforts of Henry Ford, but also to the dedicated push to create thousands of miles of new roads - the two efforts supported and fed upon each other.

What, then, are some of the industries that will be vitalized by the New Internet? One characteristic of v6 is that it will lead to millions of new identifiers and sensors (including webcams, chemical sniffers, radiation and heat sensors, and so on), connected with end-to-end access capability. Determining "who" all these entities are is being well studied, with advanced addressing schemes and allocations. But the inevitable next question that will be asked of our intrepid armies of v6-enabled entities will be, "Where are you?" The answers to this question requires new generations of advanced visualization and geographic location technologies. Let's illustrate some of those issues by looking at a couple of the players in that field - any one of which might become a strategic partner for vendors of IPv6 communications products or services.

One related area that is expressing surging growth is GIS (Geographic Information Systems), the marriage of location descriptions with related data. GIS got a major boost with the end of the Cold War, when troves of pictures from Russian spy satellites covering most of the world suddenly became available, followed by private satellite photo services. Combining photos of terrestrial features such as buildings or city blocks that were taken from satellites or aircraft at different angles enables the generation (via interpolation or "morphing" techniques) of 3D models of those features, allowing city planners, emergency responders and travel agents to move through and explore virtual representations of those areas (like most early 3D animation, this was first developed for the military). Such graphics can be combined with underlying data to provide unified three-dimensional "situation assessments" for both military and industry managers. For instance, the City Fathers of Richland County, S.C., created a detailed 3D model of the city Columbia that is used for all major county services, and can even be used by tourists to "fly through" downtown (see www.richlandmaps.com). The Israeli military decided that much of their entire country should be modeled; this was done by Tiltan Systems Engineering, which created photo-realistic models of towns so that the IDF can rehearse missions on a "digital sand table" which realistically simulates the urban areas. Medical teams can look at 3D models of a country and see where an outbreak is originating from in order to save countless lives (this is being done with cholera in Bangladesh).

The combination of 3D GIS and IPv6-enabled entities (people and sensors) is a potent one, especially with the new generation of low-cost GPS chips for position location of moving objects. A v6-enabled car reporting that it had an accident, for instance, might get a far different response if it reported "I just had a rollover in the middle of a clover-leaf highway junction" than if it relayed "I just had a fender bender in the suburbs." A DHS (Department of Homeland Security) manager could track his teams not only by which city building they were in, but on what level of the building they were located - far different than the situation today, when he might only know the approximate map location of his people, without knowing whether they were in a skyscraper or a vacant lot. Tracking people inside buildings calls for another related technology, of course - "breadcrumbs" which can be dropped in sequence when entering a building, to keep the line of communication/reporting open, similar to the line of breadcrumbs dropped in the fairy story forest of Hansel and Gretel.

There are four major players in GIS: ESRI, Intergraph, MapInfo and Autodesk, though there are many smaller companies and startups (for a sample, see www.urbansimulation.com). ESRI does close to $500 million is GIS-related sales per year, including both data collection and animation tools within the basic toolset, named ArcGIS (expect to hear this name at future IPv6 meetings). Autodesk is especially well placed in this industry: it has tools to describe large areas such as cities; it has a major CAD toolset to create models of individual buildings; and, it has a 3d animation package (named 3d studio max) that can be used to create characters and crowds and move them around in the 3D city areas - you have seen many of the creations from this software in the blockbusters at your local multiplex.

3D descriptions of areas gathered from photos by far-away aircraft is good enough for rough work, but for many other applications -- such as homeland defense operations, where teams will want to know the details of "Where's the front door on the Starbuck's?" - more detailed descriptions of buildings, roads and even trees is necessary. Occasionally the original CAD drawings of a building may be available; if pictures of the outside are available (to get the colors and textures of the exterior), the building can then be created as a 3D object and placed on a virtual street. But working with CAD drawings of hundreds or thousands of buildings is laborious and expensive, and many such drawings are no longer available. How do you get a 3D picture of a whole city block, that can then serve as a background - for instance, for illustrating where IPv6-enabled webcams are located, so that peacekeepers can activate them and look around for bad guys before moving into an area (this was done routinely for SDSU's "shadow operations" at the Super Bowl and other events)? How do you get a photorealistic digitized shape of something huge, like a football field?

What you need is an industrial-strength "scanner" that can send out millions of laser beams and form a "point cloud" of an area such as a building. One company that builds such scanners is Riegl, of Orlando, Florida. One of its scanners (which is about the size of a loaf of bread) can be mounted on top of a vehicle and driven down the middle of a city street to create 3D models of the buildings, a whole block at a time. A co-mounted digital camera provides textures and colors that are "pasted onto" the 3D models created by the scanner, to result in a final photo-realistic representation of the whole area, which can then be digitally stored and made ready for "virtual walkthroughs" by visitors.

Speaking of storage, how do you store and retrieve the massive data resulting from 3D GIS and thousands of v6 entities? Any kind of graphics is a memory hog, and 3D graphics is that in spades - Hollywood post production studios that deal in blockbuster special effects have measured storage in TBs (terabytes) for some years, and now routinely speak of PBs (petabytes), terms that are still remote for the general public. Holographic storage may be a fitting solution for putting away all the data that IPv6 networks are likely to generate - as well as the 3D contexts for that data. Colorado-based InPhase is a promising example of a company working on optical storage. Its initial product, the 200-R, can store 200 gigabytes of data in a cartridge that looks like a DVD - only somewhat thicker - with a transfer rate of 160Mbps. The storage capacity is expected to exceed 1.2 TBs by the end of the decade. Optical storage works by splitting a laser beam into two parts - one that carries the data and one that is the reference beam. Where these two beams intersect in the light-sensitive recording medium (which in the case of the 200-R is a type of "glue" between two thin discs) forms the holograph. Conventional storage records one bit at a time - in the case of a DVD, a location on a 2D surface either has a pit (a "1") or not (a "0"). In the case of the 200-R, a picture is generated of a "page" of data - a checkerboard pattern of a million zeroes and ones all at one time, and this pictured page is then stored in a location, rather than just an individual bit. Changing the reference beam allows many pages to be stored in the same location. The optical storage medium is 3-dimensional, not 2-dimensional as today's magnetic tape or DVDs or hard drives are, providing for many layers of storage locations. Finally, search rates are very high, and search can be by "patterns" instead of just by addresses, so a future query might be, "Look for a network of sensors that looks more or less like this," instead of "Look up storage address XYZ123."

In short, it sounds like interaction between major IPv6 companies and related technology firms that can help promulgate v6 products should be high on the agenda of those that want to dominate future v6 markets. Japanese companies have already been doing this; they have some advantages, as many of the ancillary technologies can be found under one roof - Mitsubishi, for instance, has major resources in advanced displays, sensors, storage technology, wireless networking, and even protected military vehicles within which to "package" mobile applications of IPv6 management centers. It may be time for companies in America and elsewhere to have that same "big picture" outlook, and reap the benefits of successful strategic teaming.