EDS' Next Big Thing Blog: Read and Respond to What the EDS Fellows Say About Technology

Microsoft's New Surface

Microsoft’s new table top computer may be more than just another PC. Because of the unique capabilities of its “touch screen,” it may represent an early step along the way to more user friendly PC interfaces. With its “multiple simultaneous touch screen,” it will allow users to do things that the old one-at-a-time touch screens couldn’t. Though pricing is still at the “very early adopter” point, it will surely become affordable within the next few years; when it catches on. If you doubt that it will catch on, just watch the videos.

Is this a shot across Apple’s bow?

As Planned, iTunes Without Copy Protection Have Arrived!

In a previous blog article, I discussed the EMI/Apple deal to put unprotected EMI songs on iTunes. Well, the time has come!

A new version of iTunes is being pushed to my Apple Mac as I write this. It appears to be a necessary first step required to support EMI’s promised unprotected higher resolution music. The upgrade went quickly and uneventfully.

Running the new version of iTunes, I find that the iTunes store now has an item called iTunes Plus; it’s where you go to get these new unprotected higher resolution songs, but before you are allowed entry, you must accept a new licensing agreement; no surprise there.

Once in the iTunes Plus area you have the ability to purchase many unprotected albums and songs. As expected, there is limited availability but there are some definite old favorites of mine. Thirsty for more information about when more iTunes Plus songs will be available, I click the “read our FAQs” link, but I receive no real information about what's to come, or when.

There is additional information in this story that I found on MSNBC’s Web site, but not much. Only time will tell, but this could be the beginning of a very good thing.

The Superconducting Power Grid

This story about a more robust power grid for Manhattan may lay the groundwork for better power delivery systems in other large cities. The use of superconducting cables to connect the cities power substations is a first that will be bankrolled with $39 million by Consolidated Edison and the U.S. Department of Homeland security.

I do wonder about the whole cooling problem. Even though they are using high temperature superconducting materials, they still will need to keep them cooled to -298 degrees Fahrenheit.

If it all works out, the benefit could be a more resilient grid for Manhattan and a bold step into the next generation for redundant power transmission systems in general.

Jumpstarting Computer Science In Academia

Computer science enrollments in higher education peaked in the 1980s. Since then computer science, as a college major, remained popular until a period of rapid decline began as we entered the new millennium. According to the Computing Research Association, a 70% drop in freshmen indicating computer science as a probable major has occurred since the year 2000. Such a decline is probably not an anomaly and therefore indicates a troubling trend for computer science departments in academia.

This article discusses how some colleges are retooling their programming classes and curriculum in a way that will help re-kindle interest in computer science as a major. The article discusses a number of approaches to spark interest among students. There is an underlying assumption that a big part of the declining interest exists because colleges have previously taken all the fun out of programming. Something the article identifies as the "prime number" syndrome. As a result, one of the techniques being used to increase student interest is to make the study of programming more fun; whether this single pronged attack will actually work remains to be seen.

Computer science is foundational to so many modern disciplines that I can’t help but wonder whether we are making a mountain out of a molehill when it comes to worrying about how many computer science majors we are producing. Many of the most successful scientists, engineers and designers use computer science as a tool to further their achievements in disciplines outside of computer science. Perhaps computer science makes more sense as the minor degree in these cases.

Don't get me wrong, I think we do need to have people with degrees and advanced degrees in computer science, and we need lots of computer science departments in colleges and universities that can focus on the advancement of the science. But most programmers don't need to be computer scientists.

With many of our computer programming, design and architecture jobs in information technology (which I always think of as being on the business side of computer science); we may not need that many computer scientists after all. In my experience, information technology has traditionally been populated by people from all disciplines with many different degrees and levels of education (including a minority of computer scientists). What they all tend to have in common is an interest in computers.

From my perspective, the current ideal candidate for a successful information technology career would have lots of business and industry knowledge, be computer savvy and would sport an MBA.

A Trickle of RFID Tags

The IT business has been talking up RFID tags for years now. They keep getting smaller and cheaper as they enable a number of expanding capabilities. Initially they were adopted by the security industry and used primarily for access control. For years now, many of us have used RFIDs embedded in ID cards to gain entrance to our workplace. RFIDs have given us car doors that unlock when we touch them, toll collection systems without tollgates and gas pumps that bill us with the wave of a bauble on our key chain. Even with all this, adoption has been slower than I have expected. This article asks the privacy question; is RFID a brilliant idea for business but a lousy idea for the rest of us? Perhaps such questions, by there very existence, constitute an inhibiting factor.

Taking inventory of my own personal RFID footprint, I find that I have a number of tags that surround me. There is a tag velcroed to my windshield that automatically pays tolls as I pass through the tollgate, another one that allows me to start my car (which uses a button rather than an ignition key) and one that gives me access to my place of business. There is also one implanted in my dog. There may be others in my purview but those are the only ones of which I am currently aware. It is interesting that my microprocessor footprint is at least an order of magnitude larger than my RFID footprint. Ultimately, I expect that to be reversed.

This BBC News article shows just how small RFIDs are becoming. In today’s landscape, these tiniest of chips seem like a solution looking for a problem. The company that created them, Hitachi, is studying them in order to determine possible usage. There are currently no plans to commercially produce them, but I am guessing that it’s just a matter of time before increasing RFID usage significantly raises demand. There is also the possibility that an unexpected RFID intensive solution will suddenly take off, one that will require huge quantities of chips; like using them as telemetric markers or embedding them in currency.

A few RFID product areas of recent interest include using RFIDs for people tracking, baggage tracking, payment systems and access control. These should increase the demand for chips and help drive the price per chip below the current level to an average of 5 cents per chip (often claimed to be the magical number that will spur mass adoption). These are also good examples of how current “normal sized” RFIDs are suitable for most of today's solutions. The demand for very tiny and cheaper RFIDs may not ratchet up until the wholesale usage of RFIDs for things like product tracking and identification become the standard. How about an RFID tag stuck on every product in the marketplace? I guess Hitachi's tiny chips could ultimately have the tiny price necessary to make that happen, but then there's that privacy thing ...

Action as a Service?!

As I think about the state of collaboration, document management, workflow and social computing, I believe we are at a cross roads. Software as a Service can move from an enhanced ASP model to a real driver of organizational improvement.

According to Forrester, 51% of organizations with 20,000 or more employees have a centralized IT organization. This means that many organizations have a very decentralized view of computing. Much of their collaboration needs are being met by ad-hoc solutions from collaboration service providers (e.g., Google, Microsoft...), if their being met at all. Some individuals doubt that SaaS is real, but this is one space where there seems to be a great deal of activity.

As the capabilities of social computing are extended into the business environment and the use of pattern recognition and workflow techniques matures, it is only a matter of time before Action as a Service becomes a reality.

This is not just action within an enterprise, but a federated approach where alliance partnerships can define a goal and orchestrate support, tackling some of the tougher issues beyond the capability of any entity alone. This is one of the drivers for EDS' Agility Alliance.

Security federation is now mature enough to support the activity. The tools are out there to enable to effort. The only thing that remains to be addressed is the organizational insight and desire to make it real. We're already seeing that within some organizations, the question remains when it will be offered as a service to the market in general.

A whole new meaning to the concept of Rat Race

I stumbled on this article in the BBC news about an office workstation/treadmill. The Mayo Clinic has developed an office workstation that they claim should enable workers to lose 30kg a year by enabling them to walk while they work.

It got me thinking about the range of possibilities from power generation for the office to the effect it would have on the length of conference calls. Assuming that you wouldn't be hoofing it during a call, I could see someone drawing out the call just a little bit to prevent getting back on the hamster wheel.

I have talked to some folks who have a stand-up office configuration. It's something that has always had some appeal, since it does limit meeting length and improve personal energy usage, without the gimmicky hardware.

When The Internet Is The Boss

In a previous blog article that I posted on this site, I discussed Amazon’s Mechanical Turk. At that time, there were a few examples that I used to explain the point of it, but I don’t think that they were particularly compelling. To me, it was more of a curious science project than a practical option.

Today, while surfing my usual sites, I encountered a much better example; the People-Powered Search. Although the search effort for Jim Gray and his 40 foot yacht was not ultimately successful, it is clear to me the real world potential for tools like Amazon’s Mechanical Turk to succeed in brokering such initiatives is not far fetched. With the Internet’s ability to engage vast numbers of willing netizens in real-time, such massive efforts could be quickly undertaken and completed.

One of the lessons of this story is that the concept of using the Internet as a tool to bring the superior visual skills of human brains to bear on particularly thorny "needle in a haystack" visual problems (that are beyond the current capabilities of computers) may soon become a practical mater. It is an approach that places us, as willing participants, at the disposal of the Internet.

The People Powered Search article points out that, besides Amazon’s Mechanical Turk, other players are entering the people assisted Internet game. Some of their examples include ChaCha, Polar Rose and PreFound.

Visual skills are only the beginning; as computers approach human levels of visual skill, opportunities for other superior human skills will likely surface. This "Mechanical Turk" concept may be with us for a very long time.

The NanoMetro Mashup

If you've ever wondered just how much nanotechnology activity is going on in the United States, then this article may point you to just what you are looking for. The "NanoMetro Mashup" map shows where nanotechnology activity is concentrated and allows you to zoom in for details like company names and Web sites. It also uses "pins" of different colors to break the activity down into eight different categories, like electronics, materials and research. Zoom in and check out your own location, you may be surprised at what you find.

On a related note, isn't Google Maps great?

More On Motion Sensing Interfaces

The runaway success of Nintendo’s Wii is probably a result of its motion sensing game controllers combined with its low price. One of the benefits to motion sensing in a game console is that it makes casual gaming easier for many of us who choose not to develop controller button expertise but have no problem with going through the physical motions necessary to play a game. It turns out that, for many games, adding motion sensing to a game controller reduces the learning curve to the point where the occasional casual player can, almost instantly, become adept.

In a recent blog article I mused about how these motion sensors would find their way into all manner of things, perhaps even leading to a practical "gesture interface." In another blog article, I wrote about motion detection user interfaces for mobile phones, now in production. So, while motion sensing is being successfully adapted to support some interface functionality for mobile phones, motion sensor based interfaces may be languishing when it comes to the more complex requirements of a mobile pc.

While a rich gesture interface is what I keep hoping for, it may be a while before anything really important happens. Could it be that we don't yet actually know enough about grafting motion detection onto a mobile compute platform's interface in a way that would yield useful integrated gesture based interaction? British Telecom is taking up the challenge but, as this article states, we may still be a long way from understanding exactly how such an interface would look and work. The article compared our current understanding of motion sensing interfaces with our understanding of how to use the computer mouse way back in the 1970s. Although I hadn’t thought about it that way before, I now think they may have a good point.

What the infant’s brain tells the CIO

Could you imagine the brain of an infant telling us something about future information systems?

I could. Consider the findings of the following piece of research. Between the age of 6 and 12 months, we learn the phonetic units of language. Studies show that infants are capable of discerning differences among the phonetic units (sound elements) of all languages, including native- and foreign-language sounds.

In a first experiment, 9-month-old American infants were exposed to native mandarin Chinese speakers in 12 sessions over a 4-week period (a control group also participated in 12 language sessions, but heard only English). During these sessions, native speakers of Mandarin read from children's books for 10 minutes and played with toys for 15 minutes. Then learning effects such as speech discrimination were measured, showing significant learning.

In a second experiment, infants were exposed to the same foreign-language speakers and materials, but now via audio-visual or audio-only recordings. Guess what happened? Nothing! Foreign-language exposure in the absence of a live person shows no learning!

Seemingly, phonetic learning from complex language input relies on more than raw auditory sensory information. It needs referential information, environmental information, social cues and context.

This reminds me of the "ubiquitous edge," the anticipated growth of sensors and peripheral "devices (such as RFID), and the associated exponential growth of content and context information.

In other words, the emerging context awareness of devices and the multi-channel, correlated input feeding our information systems might boost the learning and adaptation capabilities of future IT systems and consequently take the agility of organisations to the next level.

Sun Spots

In a presentation we had from Sun the other day, they were talking about the Sun Spot. This appears to be a very versatile and interesting device that can be used to enable various sensor and proximity based Proof of Concepts. The cost today is a bit high for embedding in anything, but as a tool to demonstrate technology it looks fascinating.

The device that came out of Sun's research lab has the following features today:

Hardware:

• A Sun SPOT device is built by stacking a Sun SPOT processor board with a sensor board and battery
• Sun SPOT Processor Board
• 180 MHz 32 bit ARM920T core - 512K RAM/4M Flash
• 2.4 GHz IEEE 802.15.4 radio with integrated antenna
• USB interface
• 3.7V rechargeable 720 mAh lithium-ion battery
• 32 uA deep sleep mode
• General Purpose Sensor Board
• 2G/6G 3-axis accelerometer
• Temperature sensor
• Light sensor
• 8 tri-color LEDs
• 6 analog inputs
• 2 momentary switches
• 5 general purpose I/O pins and 4 high current output pins

Software:

• Squawk Virtual Machine
• Fully capable J2ME CLDC 1.1 Java VM with OS functionality
• VM executes directly out of flash memory
• Device drivers written in Java
• Automatic battery management
• Developer Tools
• Use standard IDEs. e.g. NetBeans, to create Java code
• Integrates with J2SE applications
• Sun SPOT wired via USB to a computer acts as a base-station

It looks like a very interesting opportunity to demonstrate some fairly sophisticated concepts.

Now if it only worked with Microsoft Robotics Studio as well...

JavaFX For Rich Mobile Interfaces

On May 8th, Sun Microsystems announced Java FX, a strategy and products to both improve the uniformity and simplify the development of content rich Java based Web applications on a wide variety of platforms. JavaFX Script and JavaFX Mobile are just the beginning, other JavaFX products are intended to follow.

To understand the potential impact of JavaFX Script, one need only look at competing products. The most significant ones that come to mind are Adobe Flash and Microsoft Silverlight (currently in beta). Sun must be thinking that the time is now right for Java to enter the fray.

So when we think about the kind of media rich interfaces that JavaFX Script will turn out, we should be thinking easy to write, Java ecosystem, mobile platforms and animation. It will be considered an added benefit by many that it is open source as well.

Self Assembling Micro Chips

The concept of nanoscale self assembly has been one of the mechanisms anticipated to be important for certain kinds of nanotechnology based manufacturing techniques. It looks like that concept is becoming a reality. The novel approach outlined in this article is characterized by a manufacturing technique that "grows" insulating air gaps as part of the microchip production process. The technique is compatible with existing manufacturing materials and facilities, and the resulting chips are 35% faster while using 15% less energy when compared to similarly scaled chips that use traditional insulators. Such self assembled insulators could show up in production chips as soon as 2009.

A related BBC Article likens this self assembling of air gap insulators to natural processes that mimic Mother Nature's method of producing the light and airy snowflake. More proof that proven processes (like Mother Nature’s best practices) are often worth leveraging. In this particular case, we get both a nanotechnology breakthrough and another positive increment for microprocessor performance.

Mobility Comes To Linux

Is it necessary for desktop operating systems to have associated miniaturized mobile versions? Microsoft Windows has Windows Mobile, Mac OSX will have its mobile variant in the new Apple iPhone and now, it appears that Linux could have a serious mobile solution as well.

An Ubuntu Linux project is expected to release its initial version of mobile Linux in October of this year. As an open source mobile operating system, mobile Linux’s zero cost sets it apart from proprietary solutions for both Windows and OSX. Unfortunately, that still may not be enough to guarantee its ultimate success.

The good news is that a partnership between Ubuntu and Intel is behind this mobile Linux activity. The mere existence of an Intel partnership serves to add credibility to it while a recent demonstration, by Intel, of a device running the mobile Linux prototype adds weight to the story.

Ultimately, if successful, a mobile Linux could have a significant impact on future cell phones and other handheld devices.

How much are networks like environments in nature?

I was listening to the Scientific American podcast the other day and there was a story about parasites and their role in determining the health of an environment. One of the statements that was shocking to me was that there are 10 times as many other organisms in the human body as there are cells. You can tell if a higher organism or any environment is healthy if there is a diverse and healthy set of parasites living in it.

It got me wondering about corporate networks and moving from private networks to having more and more business traffic going across the Internet. Is there an analogy here to the parasite model for organisms? Denial of service traffic is similar to when an organism gets out of control and overloads the environment. Cancer is a similar situation within an organism. Are there examples of how nature copes with parasites that we should be looking at for having a healthy corporate network, that we're missing?

Nature uses a variety of techniques to handle infection, since any single solution is eventually overcome, much like we're seeing in responding to network security concerns.

Fixing The Virtual Keyboard

One of the problems with virtual keyboards is that you don’t get the kind of tactile feedback that enables reliable touch typing. That little click that you feel through your finger tips when you press even the tiniest of mechanical keys makes all the difference in the world when it comes to reliable touch typing. To substitute for this lack of tactile feedback, most virtual keyboards offer an auditory click. In my personal experience this click helps but it is not enough.

This MIT Technology Review article discusses research currently underway to solve the virtual keyboard touch typing problem using techniques currently emerging in haptics research. The research presses the idea that specific vibrations could be used to generate feedback each time a key is touched. According to the article, there are Samsung mobile phones that leverage the built-in vibrate hardware to accomplish this already. This crude approach is only the beginning, advancements in haptics and actuators will continue to refine this approach. I think I would still miss the sensation of key travel but we’ll see.

Taking this idea to an extreme, I can’t help but speculate on what we may someday see; a touch screen that applies complex tactile feedback to any location on the screen so that it could be selectively applied directly to the location that is being touched. With this kind of “tactile resolution,” when a finger misses and lands between two virtual keys, the sensation of each key edge could be applied to the offending finger. Now that would be a haptic touchscreen.

For those of us awaiting the iPhone, the above article from the MIT Technology Review opens with the disappointing revelation that there probably won’t be any tactile feedback on iPhone's virtual keyboard. Based on this article from the New York Times, that revelation appears to be true, but we probably won’t have the final answer until production units are in hand.

Enterprise News of the Future

I came across this blog entry about TV News of the future? and instantly thought about how organizations could use a similar technique for their executives to be informed about events within their enterprise or industry.

You could add attention engineering techniques to say which of the news items should go to the mobile device ... It doesn't have to be limited to just news feeds or corporate press releases, you could do something similar with workflow or ERP system information ... but then we're just getting into the realm already possible with executive dashboards.

When I think of information overload, I think of it as bad context or personalization, maybe something like this would help.

Multi-processor Programming Languages

Now that multi-core is all the rage, the daunting task of writing programs that take full advantage of multi-processor power has thrust itself upon us. We’ve talked about this problem for years, and there has been some progress toward a solution, but now that multi-core is in rapid adoption, the issue of multi-processor programming is beginning to loom large.

The ideal solution is a programming language that, through automatic internal machinations of both the compiler and the resulting execution code, creates implicitly parallel programs. With implicit parallelism, a programmer need not be concerned about how many processors a program’s target computer has. The responsibility for slicing processes into parallel chunks is one of the duties of the language (whether it be handled by the compiler or the resulting execution code or both is irrelevant to the programmer). As with any ideal, this concept may fall short in practice.

Researchers at MIT have been looking at this problem with an eye on easing the pain of parallel programming while creating a language that generates efficient and error free implicitly parallel programs. The resulting language, called “StreamIt” is in a class called “dataflow languages.” Although not a new concept, the article implies that StreamIt creates parallelism via a data flow architecture that enables the programmer to program “normally.”

Because implicit parallelism is already available in languages like ZPL and SISAL, I can’t help but wonder whether StreamIt is better or just different.

SOA still siloed after all these years

I came across an article in SD times titled Will SOA Become the new Siloed App? that's well worth the read. It reminded me that today's technical zealots are tomorrow's Luddites. We need to reassess our views when we use new tools.

One of the things I say to people thinking about going down the SOA path is that "there is no such thing as a departmental service." Once the service is used in the corporation, it needs to be treated as an enterprise service.

Although SOA (by definition) is silo based, the silo needs to be around the service and its role in the business process, not around the department or a manager's perspective. Many times the projects will be initiated within a department, but the enterprise rules and governance should be followed. Act locally, think globally. There are some significant concerns that arise related to how services are paid for, maintained, updated and tested that go far beyond the needs of a single department as soon as the services are picked up by others.

One of the items stated in the story is that SOA is rarely initiated at an enterprise level and it "typically occurs only when a company is in trouble." Since this is true, most people will encounter it at the departmental level first, but some level of enterprise rules need to be in place.

Speech recognition in Vista

Last week, I finally switched over to a Fujitsu tablet PC, Office 2007 and Vista for my main work machine. It took a little more time than when I switched my home machine over to Vista last fall. I do all kinds of crazy things at boot time to start up my work environment, based on what resources I can see ... naturally, I had to move all that startup work over to PowerShell. I am relatively happy although there still seem to be some lingering issues with drivers (e.g., my Creative desktop camera).

The Vista handwriting recognition is orders of magnitude better than when I did some pilot work with Microsoft on their Acer prototype tablets a number of years ago. My writing is not perfect, by any means. In fact, sometimes I wonder how it is able to recognize what I wrote, since I was having doubts. I’d say it is in the high 90% accuracy, depending on what mode I’m trying to use.

The one area that still seems to have a long way to go is Vista speech recognition. I was going to “write” this blog entry using speech recognition, instead I am writing the entry about it. I must be doing something wrong, since it still doesn't seem to recognize worth a darn. I spent over an hour running through training (more than once) and I've yet to see it get an entire sentence correct. I don't think my speech had any noticeable accent -- I know, you've probably heard that before -- but I grew up in Indiana. News anchors get sent to Indiana to learn how to remove their accent.

A number of years ago, I played with the Dragon speech recognition software and it seemed to do a much better job with no training, even though my computer was probably a third the power of this one. I am hoping I just have not configured something right yet. The problem is that most people are not going to give it more than a second chance. I may need to dust off my really old Dragon software and see how it works under Vista, since I have written some EDS internal blogs with it back in early '05.

Next I’ll switch my recreational programming over to Orcas and see what comes of that.