This is an experiment. After coming back from Consumer Electronics Show in Las Vegas. I'm trying to distill what I've seen. I'm not sure the audience for this debrief should be -- feedback would help. One caveat is that this is not a polished document. Just a collection of thoughts and observations.

For those not family with it, the http://www.cesweb.org is sponsored by the Electronics Industry Association (EIA) and, in particular, their CEMA (Consumer Electronics Manufacturer's Association). The show itself was dying two years ago but has come back strongly (though the summer show has been dropped). In fact, it seems to be eclipsing Comdex as the computer technologies become mainstream. This was evident by the larger CES airport traffic as compared with Comdex!

There are still some parts of the show which are definitely not Comdex such as booth selling watches from Taiwan and the Adult area. Though the letter is getting more web-oriented than being focused on video and accessories.

My particular interest is in the IP Infrastructure and in how connectivity is affecting products. Part of this is my "enlightened self-interest" agenda in pushing the IP Infrastructure concepts. Of course, I'm looking for products that I would like to buy. These overlap in current focus is on devices that enable me, as an individual and hobbyist, to build things on my own.

Part of the show is organized into "Pavilions" or groupings by theme. I'll use these loosely as a focus.

Pavilions

Digital Living Room

This is sponsored by Upside magazine. The comments here are not strictly limited to exhibits in this Pavilion.

• HomePNA. This is the Home Phoneline Networking Association. This is particularly close to my heart since I created this technology. More recently Epigram has moved it from 1 mbps to 10 mbps! This actually works and, combined with Microsoft's Internet Connection Sharing (ICS), I have been helping mortals network their homes by just connecting their PC to via phone line and clicking on ICS.
  • Compaq ships its own alternative to ICS. The advantage is that, unlike ICS, it supports multiple networks so one can have a house with CAT-5, phone wire and wireless. Acer bundles Sygate (since they had to have a solution before ICS), which supposedly gives more control. But, alas, they haven't tested what happens when a user with an ICS machine then install Sygate. This is a problem endemic to the industry – companies that ship solutions that work in a limited number of cases but they do not take a wider view and end up causes damage outside their narrow world. Diamond Multimedia suffers from the same problem.
  • Epigram, as I mentioned, is moving ahead to higher speed and can run DVD streams over their network! More interesting is the ability to use the phone wire technology as an alternative to USB. It can be as cheap and, supposedly, with Microsoft's UPNP (Universal Plug and Play) one can just plug peripherals onto the Ethernet. This story needs work and I will need to research UPNP as well as Jini but it is a stealth alternative to USB. I've been campaigning for using Ethernet (IP) instead of USB but my take is that it will simply happen as we make it work and the limitations of USB become evident.
• HomeRF. Well, it's still not shipping but Compaq was demoing it and Bill Manny of Compaq is still at it. Intel was also demoing an ADSL gateway with RF connectivity. But it faces competition from 802.11. HomeRF is supposedly lower power and cheaper but 802.11 goes faster. My take is 802.11 will win because HomePNA will be the low price alternative thus reducing the need for a low cost RF option. Bluetooth (see below) is trying to be a spoiler. Even though its range is limited, it confuses the issues. On the other hand, Siemens seems to have announced their HomeRF support. 
• Phones.
  • This is a bit a field but I'm looking for a replacement for my Siemens Gigaset because the Gigaset is decaying and Siemens doesn't have the concept of software upgrades. But Siemens was there so I may call the support people anyway. Siemens uses the US version of DECT (Digital PBX protocol). DECT is problematic in that it is an ITU standard with its limitations built in.
  • Home Wireless Networks in Atlanta was showing their phone system. They do both data and phone but limit the total handsets and computers to 8 (or 16 for the office system, but at a premium price). It looked more interesting at the show. Looking at the web site it seems to aim too low – 900 Mhz star configuration with an upper limit on the number of units and is oriented towards dialup (including ISDN) connections rather than higher speed connections.
  • CyberGenie seems to have better design point than Siemens. It works as a simple cordless phone system on its own but can use the PC for enhanced service. Alas, it's "Phone Assistant" is too user-fawning and can be annoying in trying to do voice recognition and services but, perhaps, it can be ignored. The added services do require a PC but that's OK since they seem noncritical. Of course, the trend of having a simulated human assistant would have been a disaster if it were available when ATMs were first created – the ATMs would have tried to preserve all the worst properties of human tellers. A nice feature of Cybergenie is that they use small Ericsson handsets. They also see themselves as a software company that does upgrades. Cybergenie is also USA-DECT.
  • Panasonic has home office PBXs and was showing wireless systems but I didn't get a chance to check them out but they may be interesting.
  • TT Systems is selling their phones as IBM. I've used their distributed key system in the past and, again, might be interesting in the cordless arena but, again, didn't look in full detail.

Digital Hollywood

Of course, the entertainment industry is discovering the Internet et al. Again, I'm including other "media" stuff here, not just the Hollywood stuff.

• Packet Video. Interesting in that they are doing streaming MPEG-4 over low speed connections. The problem is that they seem to be trying to do this by making deals with the companies in the middle such as Telcos and buying into the WAP world. They don't understand that they can be an end-to-end play and sell a product directly. Perhaps I can find a useful way to help them.
• MGI is doing software to make PC's be the basis for home entertainment and TV systems. But they are only selling it bundled and in Asia. I'm interested because Tivo and Replay annoy me since they should simply be software apps. NetTV n the US is working with them and will be producing (on their second try) a US offering for a PC that can be the heart of an entertainment system. As an aside, Microsoft was doing some of this a while ago but, well, blew it. But that's another story. Overall, however, the PC is going to become the entertainment vehicle since there are a number of companies doing HDTV decoders on their video boards and companies are doing HDTV decoding in software. I expect some of this to gel within the coming year. ATI, for example, has such capabilities in their current boards but it doesn't work as well as it should. Rivisent is partially owned by ATI and is doing new software.
• BroadBand Magic sells computers that act sort of like Set Top Boxes. In particular, http://www.broadbandmagic.com/computerplus_spec.htm.
• Digitra supplies HDTV boards for PC's. They happen to also have a line of MP3 products.
• Digiscents is doing what is, well, Internet-based smell-o-vision.
• SpikeRadio is an Internet-Radio company. What makes them interesting is that they are working with Toshiba to bring IP Radio to the car. I don't know the details but I want to encourage any effort to bring IP connectivity to cars. Once that happens then RAWKI (Radio As We Know It) ceases to exist.
• SonicBox sells an IP radio that acts like a traditional radio including having a tuner. There are PC-based products and a separate box that connects via a PC though apparently others have Ethernet connectors. They also sell a board that can be embedded in a clock radio or other device. The board includes an IP/Ethernet interface. One nice aspect is that they dynamically determine the traffic flow, unlike the current streaming players, which ask the user to give a number – an idea that doesn't work in the real world. These devices emphasize the point that any UI advantages of RAWKI over IP radio are simply artifacts of current products. This kind of AWKI defense of radio, for telephony or just about anything else is simply shortsighted.
• InetCam is interesting in that it packages Webcams as consumer technology. Its software covers a wide variety of cases. It can automatically stream to a web site (including free home pages). It also will provide you with a dynamic DNS entry – your machine will update its current address when it connects to the net. This is not a new idea but the value of INetCam is in making it all "just work" with a camera you plug into the PC. Whether this is the right product or not is an open issue but, once again, it does demonstrate the ability to package solutions at the end points. It also demonstrates the value of the PC as a platform for innovation even as I lament its complexity and look for interesting devices. In fact, these devices are simplified since they don't have to duplicate all of the capabilities of the PC and can focus narrowly on applications.

The Home

This isn't really a pavilion but since I'm using it as an organizational theme … some things such as phones have been covered above but there's lots more.

Home Connectivity

Already mentioned HomePNA and HomeRF above. A couple of companies are trying to build on "broadband" connections to the home. Intel and others over gatewaying boxes. 

• CoActive is trying to provide connections for Echelon and other technologies. They are interesting because they are emphasizing what one can do with a network connection in terms of IP-based services.
• 2wire is a "residential gateway: effort that tries to also provide "voice over DSL" which violates my IP layering principles but is exciting to the telephone industry that is trying to extract additional ISP revenue. They distinguish between a residential gateway and a set top box. Their definition of a residential gateway is still a telco-oriented view in that it does house other telephony services such as VoDSL. This is similar to the set-top-box/cable view that the box also has magic decoders for encrypted video streams. The good news is that they comment on Internet Appliances that have RJ-45 or HomePNA RJ-11 connectors. (Hmm, this HomePNA stuff really is catching on!). You can look at their Residential Gateway FAQ.
• Belkin is selling their home wiring scheme that I mention below.

Home Theater et al and HDTV

This is a perennial at CES with speaker companies making their year's revenue. Still there with lots of DLP TVs and fancy screens. It's complementary to all the high end display devices being shown on the floor. But there is still a gap between these devices and what is available in retail in the US though 16:9 TVs are selling elsewhere in the world.

As noted, HDTV in the PC is much easier than TV. Midlevel HDTV resolution is 1024x780i (interlaced) which is considered low for PC's Admittedly large displays are expensive but home theater seems to depend more on sound. All this makes the convergence of PC and TV displays more obvious and one can think of a PC (either general purpose or special) as a software-defined controller. My personal view is that this will allow a development cycle much faster than traditional CE companies can match. Should be very interesting.

Even high resolution HDTV (1280x1028p?) doesn't seem all that high (I may have those numbers a bit wrong). What is really surprising is that this is the resolution used for large theater digital movies that are viewed as better images than 70mm movies! One reason is that film has a lot of jitter due to mechanical factors. Hmm…

Also, I've heard conflicting stories but my personal take is that developments in PC-based HDTV also make the PC a very good DVD decoder – better than standard players. We wind up just quibbling over monitor size. Though some claim that optics don't follow Moore's law, I've heard the same claim about printers, about telecommunications etc. The ability to evolve the components independent is the key to Moore's law. The cost of the electronics doesn't vary with screen since, only with resolution, which also abets large image viewing. Watching a video game with high performance graphics on a $30,000 large projection TV is a much more exciting experience than the same old tired HDTV demos.

Hughes makes the receivers for DSS satellite TVs and has an HDTV version. I suspect that this will be the primary deliver for HDTV since local cable companies will find it hard to justify as well over-the-air companies. But the satellite systems use up lots of channels to send the same movie shifted by a few minutes. This can be better done by buffering on disk. This would means that they have the capacity for HDTV and benefit by having a distinct advantage that allows them to skim high end users. The receiver has RF-remote control so it can be placed in the basement. Matching that with devices from Channel-Plus for video distribution allows one to shift from seeing this as per-TV boxes to simply source streams. One can decide how many streams are available concurrently and from which sources. Once again, this is a straightforward idea that hasn't quite made into the CE culture. Again, opportunity! In fact, as noted above, even HomePNA is fast enough for a video stream or two and CAT-5 at 100mbps has lots of capacity. Hmm.

Home Control

Another perennial. CEDIA (Consumer Electronics Dealers and Installers Association) is the real show for this area and this year I saw less of it at CES. But Smart is finally going to be shipping CEBus components. The advantage is that these are much more reliable than X-10. But it is a transient technology and they do agree with me (well, I've been telling them long enough) that IP-devices will have the advantage. There's a real opportunity to work with them and others. GE is sponsoring the smart device so I do expect to see them in retail and they can give those of us the ability to play with them while waiting for the better devices.

Home Wiring

Belkin is pushing their home wiring scheme but the focus seems to be on facilitating traditional wiring such as RJ-6 for video. One cute aspects are the hallow baseboards which make it easy to wire existing homes without tearing up all of the walls.

Digital Imaging

Of course, digital cameras continue to advance. I've already decided that film cameras need to justify their difficulties. While they do have some imaging advantages, for the moment, over digital, it's not enough for most purposes any more. Professionals can be high-end digital cameras that are very good. But the 3.3 megapixel cameras are about to debut!

Of particular interest was the Sanyo VPC-SX500 that does 640x480 video capture (4x my current Sharp MPEG camera) as well as 1.5 megapixel still in rapid sequence mode. Hmm, looking at their brochure, it says 1/16 VGA mode for full motion capture, which is a lot less interesting. It allows one to use the IBM CF 340MB drive for 50 minutes. Hmm, maybe not as exciting but will need to look at it some more. The camera is about $500 with 8MB CF. The IBM drive is $500.

Sanyo was pushing the ability to go from the camera directly to CDR. In the US I believe that the PC will remain the focus rather than a plethora of special purpose devices. I do believe in specialized devices where form is function but only after the market can explore approaches. In Japan, however, devices like Word Processors sell. I can speculate why but that's for another essay.

Spoke to Lexar and SanDisk to learn about the various digital media:

• CF (Compact Flash) has a processor inside. This allowed Lexar to add USB capabilities to each CF so they can do a simple USB adapter that they simply package with the CF ship. Makes connection to a PC very simple. They also claim to have patented many techniques for fast CF writing. The capacity is up to 225MB (though that is very expensive), which means that they can challenge the IBM micro drive. They also pointed out serious heat problems with that drive which makes them problematic in many of the cameras.
• SmartMedia is the form that has no processing – it's the dumb media. The advantage is a small form factor. But without processing, the transition form 8 to 16MB left many of the original readers unable to read the new 16MB media. 16MB and 32MB seem to work fine but 64MB is just coming out and is an unknown. I'm currently using SmartMedia devices but the limitations are an issue.
• Sony has its memory stick but it's not clear how much wide support it will get.
• Newest is the "Multimedia Card" which is smaller. Sanyo/Fisher is using it in their new little MP3 devices. It is smaller that SmartMedia but may be better designed for long-term improvements. It's too new to tell what will happen but give that I give it a chance since the small size allows innovative devices such as MP3 players that are more like pens than boxes..

Cars!

Cars are, as usual, a big arena for Consumer Electronics. I didn't look this year but there are normally OSHA warnings on some of the demo cars.

Digital Radio and Satellite radio. These are attempts to keep the old style radio alive with new features and quality. But, like RDBS (Radio Data Broadcasting System) which is supposed to let you scan for a particular kind of station, I expect them to fail in the face of IP radio. But, like WAP, there will be lots of investment and excitement for a period of time.

AutoPC et al

The term "AutoPC" is Clarion's version of the Microsoft WinCE platform for the car. The original "vision" was that there would be many manufacturers licensing the reference design. But, having bought one for my car, it's obvious that these companies do not understand software platforms. Instead, it is positioned as a replacement for the car radio and it takes over the DIN slot. It's a zero-sum game. I was willing to do this for my old car but my son hates the UI now that he is using it. (I did discover some short cuts so maybe it can be easier).

I had assumed it was dead but Clarion is coming out with a new model that features the ability to add video screens. When driving only the backseat is supposed to be allowed to see video in the US. (In Japan, where Video Karaoke is important, there is no such limitation). Alas, it's not an upgrade to the current unit. While I would like to get the new version, I'm loath to give up the DIN slot and junk my new car's CD Changer/Cassette Player/Radio.

Microsoft is also teaming with Visteon to do an embedded version but this is even more closed than the Clarion implementation.

So I may have to get the Magellan navigation system instead. That's the one that Hertz uses though Magellan laments Hertz' inability to install them correctly.

And I'm looking at various alternatives to put a "real" computer in the car. While one can place a laptop on the dash, an installed system with a remote screen would be nicer. One company is Real Car Computing . Their background is building systems for the military so low-price is not an emphasis.

As an aside, Pharos had a nice GPS kit for WinCE but, alas, the WinCE screens are too difficult to read in normal light.

IDB and Car Networking

I looked at the (Intelligent Transportation Systems) ITS Data Base (IDB). This year I had a chance to speak to Allan M Kirson (he is listed at www.sae.org as being at Motorola) who is considered the father of the standard. The people I spoke to last year emphasized link level reliability. Allan was much more amenable to the use of IP. The problem is that the bus started out as a protocol atop of existing car buses that are low speed serial connections emphasizing link level reliability and won't be available till the 2002 (or 2001?) model year. And the higher speed buses intended for entertainment won't follow till later.

This reminds me of the best offers of the bell-heads to do networking. One should be able to build IP networks on diffuse IR or by using the HomePNA technologies. The HomePNA design is intended for difficult environments and doesn't require a hub. Fiber and other media are options but require planning. The IDB argument is that physical medium commonality is important but, again, with IP we can bridge. Much more important is enabling innovation and experimentation.

The big problem is still IP connectivity to (and from!) the car. But a GSM phone can do 9600 bps now and soon, with GPRS, much higher speed. We are then quibbling over price but early adopters can get flat rate pricing for GSM or CDPD. The marketplace can innovate much faster than a process that is dependent upon the current large companies.

My comments about the DIN slot as the scarce resource applies here too. With a network we needn't displace the existing radio or other gear in the DIN slot.

Traffic Data

There were a few new efforts to provide traffic and navigation data. Existing players such as NavTech are aimed at commercial or other limited markets and aren't sufficient responsive to the need to iterate. For example, I didn't find navigation systems that allow me to add my own information.

Here too, just give me the data and the ability to write simple programs and, well, stay tuned!

TelConTar has their Rich Map Engine that includes Active X controls. Perhaps useful for experimenting with one's own system though I assume the pricing is commercial (high).

Car Video

In addition to Clarion's AutoPC video, there were a number of other car video offerings for the passengers. TMI Products had a standalone unit for ceiling mounting which should make installation relatively simple.

Other Notes

Bluetooth

Not much at the show. But in speaking to the Bluetooth demoer at Acer, he commented that wireless 802.11 (Ethernet) is just data (though it can do voice over IP) but then he seemed to agree (given he was technical and felt safe) that Bluetooth does give a voice path but just one. Once again, a network with its limits built in. Bluetooth also faces FAA challenges and apparently causes other RF interference in the neighborhood of a phone. I suspect that other failures of imagination will soon become evident.

WAP

Perhaps I didn't look hard but I didn't see much WAP presence. Even Sanyo, which featured their Sprint phones, didn't mention WAP. Of course there were WAP phones but it didn't seem to get special mention.

[A]DSL

Interesting to see a number of DSL efforts and little mentioned of Cable Modems. I'm not sure why. Perhaps the Telco industry is more into deal making and in taking responsibility for premise wiring than the CableCos. DSL combined with HomePNA allows a nice single box solution though the current players are only nibbling at this. Creative Labs has an aggressively priced unit at $149 that might give DSL an advantage.

Alas, DSL is still a salvage operation that requires you have the right kind of legacy copper wiring. The DSL focus is on the accidental properties of recycling phone wires with a technology developed for Interactive TV. The offering should be connectivity not a particular kind of signaling standard!

Cell Phones

I looked because I have Nokia 6100 phones. They are showing their new 7000 series (not shipping yet) that does have WAP capabilities. The good news is that it can share accessories and car mounts with the 6100's.

I didn't have time to look through the cell phone booths in details but didn't see any surprises and, as noted, surprisingly little mention of WAP. Bluetooth and WAP were more in evidence at Comdex.

How To

ETown and others are publishing "how-to" information on the net. This is another exciting trend. It's easy to understand the advertising model of a site that provides antidotes. As a user I'm happy to see sites incented to provide this information. While we can argue about the web vs. traditional publishing, books simply aren't there when you need them.

MP3 Devices

No surprise to see lots of them with a number also support SDMI (the Microsoft et al encryption standard) in addition. There are a variety of form factors and other features such as supporting Audible's news and information feeds.

Home Recording Rights Coalition

HRCC is interesting. According to their literature they want to assure that user retain the "fair use" right to record content even from encrypted sources. Gary Shapiro, the President of EIA, is involved so, for now, I'll give HRCC the benefit of the doubt in that they seem to be pursing an agenda I agree with.

Teledildonics

The idea of computer connected devices for "adult entertainment", both interactive and standalone is an old idea. FeelThe.Net was demonstrating computer-connected devices that one places on (or in?) various parts of the body. These can be connected over the network. Demonstrating them on a table, however, produces an effect more like the old "Mexican Jumping Bean". Sill, combined with the Digiscents, one can imagine the possibilities.