The New Human Frontier of Software: Mobile Computing, Page 2
The Plains of the Serengeti and Survival of the Smartest
On the plains of the Serengeti, there exist predator and prey. The relationship between zebra and lion highlights the importance of being able to sense opportunities and thwart threats. If a zebra fails to sense the proximity of the lion and thereby fails to respond by fleeing the scene, it is unlikely that the zebra will survive. Conversely, if the zebra runs away from non-genuine threats, it will deplete itself of energy and die. Similarly, if it fails to sense opportunities and act upon them it will not thrive. Real threats and real opportunities must be made distinguishable from their counterpart false positives. Dr. Chandy applies this metaphor to humans who might use historical information more intelligently to identify genuine opportunities and threats. It is the ability of software to amplify what biological organisms do instinctively that promises to bring about the vision BG presented. The emergence of complex event-oriented software systems will be an instrumental part in the realization of this vision.
Too Much Excitement
All this mobile computing excitement left me wondering about how I was going to be able to relax. This was starting to sound like serious bizzzzzness! Dr. Adam Kolawa, chairman of the board at Predixis (www.predixis.com), told me to chill. He said the whole point was to enjoy the experience; it was, after all, very personal. Predixis is the creator of MusicMagic, whose blatant marketing plug follows: "An incredible personal DJ, provides continuously varied playlists. Matching a mood or even an acoustical choice, MusicMagic can provide hours of new listening ideas. Imagine... getting up in the morning to a list of "wake-up" music. Or how 'bout creating a work-out playlist that energizes you during your cardio time and calms as you cool down? The possibilities are endless." I get a free iPod for that plug, right?
But really, I was starting to get the idea: Mobile computing will change "everything we do," including how we enjoy our music—the MP3 player is one of the killer apps for mobile computing. The great part is that the MP3 players can be used to educate children by using both audio and visual mediums. Why not learn vocabulary on your mobile device or track your scuba dives, skydives, or pilot's/sailor's logs? Why not get the surf report? I would strongly caution that you avoid being unproductive while doing these activities. Remember, the goal of computing is to free us so we can work towards greater pursuits.
This aside, I was still confused as to how MusicMagic really worked. I worried about what music Dr. Kolawa was listening to, so I asked: "Dr. Kolawa, my musical proclivities, while somewhat eclectic, are irrefragably ostentatious. Frankly speaking, I listen to gangster rap. How can MusicMagic hook me up?"
The rest went something like this:
Dr. Kolawa paused and looked at me inquisitively. "So you're a fan of Dr. Dre? Dr. Dre from Compton, California? This is good news indeed," he exclaimed. "Dre is conducting excellent research in his lab, a true pioneer in his field. You see, those Sense & Respond programs you speak of might well be able to sense the release of new songs by such gifted artists. As Dr. Dre publishes his prolific scientific contributions, we could have a sensor configured to sense this event. The Sense & Respond system would then invoke MusicMagic to analyze the song and automatically download it to you in the event it matches your 'eclectic proclivities'."
Nice theory, but I wanted more specifics. He continued, "Consider what happened with Eminem's recent release: Mosh. Within hours, thousands upon thousands had downloaded the song. Discerning Mathers fans such as yourself may have thought he fell off with that particular song [sorry Em, had to dis you]. MusicMagic might have been able to sense that the song would not be to your liking and thereby have prevented this false positive."
I was shocked, but Kolawa was relentless. "What this really means is that you'll get exposed to new artists and many old favorites. And for you, Jonathan, I believe MusicMagic will bring the joy of Nate Dogg, the smoothness of Snoop Dogg, the untamed strength of Xzibit and 50 Cent, and many more West Coast rap extraordinaires."
In one fell swoop, I felt as though Kolawa had cut through the Gordian Knot of my confusion. I was beginning to understand. There was much truth in his words. Dr. Kolawa clearly understands the importance of a highly personalized music experience, but at what price? It's comforting to know that MusicMagic is free, pre-installed on the InTempo/iPod and available for Pocket PC (now I definitely want the U2 Edition iPod, projected to be an excellent museum foyer piece in a few short years!).
But, I still wasn't getting the whole picture. What are these software systems sensing? For this, I did a little more digging and concluded it probably has something to do with the exciting research in Dr. Feng Zhao's book Wireless Sensor Networks. Dr. Zhao, the Godfather of Sensor Networks (if you would permit a histrionic reference), writes: "The technology promises to revolutionize the way we live," which seems to be the phrase du jour, a mantra in the making. It is these sensor networks that will transmit event streams containing information that can be "sensed" by the Sense & Respond Event Servers.
The Clichéd "Challenges" Section
Initially, the challenge or (to be more politically incorrect) the problem is building these complex event-oriented software systems to model the sensing of events from disparate sensor networks. There must exist a common means to define what an event entails. Does every event have a temporal element? How do you distinguish events from each other? How do you secure these events? How do you propagate events to everyone who needs to know about them? How do you version them? How do you form event hierarchies? Even though this is just scratching the surface, it is questions like these that will bridge Sensor Networks with Sense & Respond software. Once the event has been sensed, there must be a means of triggering an activity that determines how to respond. Once a course of action has been decided upon, the response must be executed. This response can take many forms; it might even include communicating back to the original sensors. This set of challenges encompasses the research at the Infospheres Group at the California Institute of Technology.
Steve Haeckel, author of the Harvard Business School Press book Adaptive Enterprise, first applied the term Sense & Respond. While Steve alone may be able to describe the adaptive enterprise, it will take the joint efforts of today's leading experts to realize the vision of tomorrow's computing landscape.
What is exciting is that Sense & Respond applications are beginning to appear. The first generation of these applications is mobile alerts. You can register your interest in certain areas of a Web site and it will alert you about them through e-mail. The coming years will witness the need for significantly more expressive event specification and subscription models to avoid the false positives and false negatives, the likes of which we have experienced with SPAM and disappointing songs. A corporation called iSpheres has created an event-oriented software infrastructure to deal with such problems. I am just ashamed of myself with these unsubstantiated plugs. I went to bed with a dog (Takoda) and woke up itching!
As an example, consider how you might apply Sense & Respond to the two prevailing branches of market analysis: fundamental/quantitative analysis and technical analysis.
You now know that the ability to react to critical conditions in near real time helps you thwart threats and respond to opportunities. Whoever is able to react quickly will have the first mover advantage and be able to take advantage of opportunities when the window of opportunity is short lived.
This first mover advantage will be instrumental for survival. For example, say a stock trades on the NASDAQ (event). This event is usually meaningless—it happens a million times a day—but occasionally it is significant. If you sense this event, you might be able to discern that a particular stock transaction is a rare event, leading to opportunity or threat. Perhaps the stock transacted event indicates that the price just crossed over its 50-day moving average (technical analysis), or the stock price broke through some resistance within a certain trading range (technical analysis). Perhaps a sensor on some rogue wireless sensor network at the Patent Office (or the FDA) indicates the approval of a company's patent on a new drug (fundamental analysis), or a sensor on the Q4 earnings reports registers a record quarter (fundamental analysis). All these events might denote a strong buy signal to institutional investors. An automated Sense & Respond software system might sense this event and respond by contacting the investor on any device (Web application, desktop application, e-mail, SMS, mobile computing device, PC telephone through voice XML, Braille?). The response would both notify the investor of the rare opportunity or threat and allow the investor to take action by placing a trade. The sensing and the subsequent response would happen anywhere, anytime, and on any device. Obviously, automated responses such as the automatic execution of stop losses also would allow traders to thwart threats as rare events indicate the presence of bearish trends.
As great as this functionality is, there is a price: Users will be more tethered to the computer than ever before. It will always be with them, their constant companion and cherished friend. So, for those rare moments when you want a little privacy, might I suggest accidentally defenestrating your mobile device, which may actually be feasible (unsubstantiated sources indicate Microsoft may be contemplating disposable operating systems).
The Proof Is in the Pizza
As a final note, consider some of the early experiences we at the Infospheres Research Group have had with mobile computing. Using Pocket PC sales automation software, we were able to cut the time required to take an order from 6:00 minutes down to 2:00 minutes. With deliverymen making upwards of 20 deliveries per day, we were able to shave well over an hour off delivery time. This meant the drivers could generate more revenue by making more deliveries. But, the efficacy improvements came largely on the backend when the driver arrived at the warehouse. We saved the driver another hour by automatically uploading all the orders and generating summaries as to amounts owed. Furthermore, we essentially automated the function of the data entry clerks and allowed management to have a closer real time view of the extended operation of the organization. Mobile computing made this possible.
For this system to truly help mankind, it must be accessible to all of mankind, irrespective of race, creed, ideology, sexual orientation, physical disabilities, and so forth. Because such a system will have an impact on the daily lives of each and every one of us, my sincere hope is that it will be used for man's betterment rather than its detriment. As we enter a new era in our own evolution, let us now use our experience and history to proceed prudently for "they stumble that run quickly" [WS—no, not Web services—William Shakespeare].
Almost done.... The technologies needed to realize the computing landscape of the future are innumerable, and the requirement for us will be bridging between theory and practice, between academia and industry. It is exciting to see such bridges taking shape in the collaboration between educational institutions and corporations, which will certainly usher in the new era that has BG saying his optimism for the future of computing has never been greater. Do you see why? Perhaps I'll elaborate further.
In the next column, I hope to divagate garrulously on the idea of the enterprise software architect and the singular software system. So, stay plugged in to developer.com (you didn't think I'd forget to plug my publishers, did you?)! And remember, nothing exceeds expectations like xCESS.
About the Author
Jonathan Lurie is a newly minted doctoral student at the California Institute of Technology, prior to which he toiled as a "professional software professional" whilst acquiring a bad case of acronym certification measles: MCT, MCSD, MCSE, MCDBA, MCAD.NET, MCSD.NET, MSF, Java Certified Developer, and IBM XML Certified Developer. He currently researches an area of Computer Science known as Sense & Respond at the Infospheres Research Group.