Waiting for the "iWatch": The Digital Watch Revisited
As we count down the days to the expected announcement of Apple’s much anticipated iWatch (or whatever it might be called), I am struck by the impact that digital watches – including their modern-day equivalents – have had on my life – which seems to consist mainly of waiting for the next cool tech gadget to be released.
When I was in elementary school in South Africa, I begged my parents to buy me the revolutionary Texas Instruments LED digital watch. They were making a trip to London, and this fabled gadget was not yet available in South Africa. This shopping expedition also served to introduce them to the pleasures of spending their money at high-street electronics stores – something that was to come in handy in later years.
The Texas Instruments could display the time – hours and minutes – and the seconds – as well as the month and date – whenever the display button was pressed. I spent a considerable part of my childhood engaged in a failed experiment to see whether time could be affected by the repeated pressing of the display button on a digital watch. The only conclusive outcome of this experiment was to establish the definitive inverse relationship between the frequency of button presses and battery life. No doubt this early research contributed to the progress that has since been made in improving battery technology.
I assume that the introduction of liquid crystal display technology was a direct response to my button-pressing addiction. Now the time was perpetually visible. No need to press any buttons. The innovation in this device was the inclusion of a stopwatch that could measure elapsed time in hundredths of a second. This led me to embark on a new experiment, which involved repeated attempts to minimize the delay between pressing the start button followed immediately by pressing the stop button. Although I managed to get this delay down to about 0.07 seconds, the manufacturer must surely be credited with the greatest achievement. There was no degradation in the electro-mechanical performance of the buttons even after I persisted with this brutal experiment for a sustained time.
Clearly I was not the only person experimenting with my digital watch. As a young post-doctoral researcher, I remember being graciously received by Prof. David Harel at the Weizman Institute in Rehovot, Israel. One of the world’s outstanding computer scientists, Prof. Harel is the inventor of the Statecharts language and methodology for the specification, modeling and design of complex reactive systems[1]. Statecharts extend conventional state machines and state diagrams by adding notions of hierarchy, concurrency and communication. A variant of Harel Statecharts was included in the UML, and so the language has found broad acceptance in the software engineering community. In a fascinating personal recollection, Harel describes how he employed Statecharts to develop the specifications for the avionics system for Israel’s Lavi fighter aircraft.[2] In this paper he explains why an avionics system is a great example of a reactive system. Yet in his formal description of Statecharts, Harel chooses to analyze the behavior of his Citizen Quartz Multi-Alarm III digital wristwatch. One of the outcomes of this exhaustive analysis is Harel’s identification of a number of anomalous states that were either incorrectly identified – or not identified at all – by the makers of the watch. These states were identified through a combination of logical analysis and the simultaneous pressing of strange button combinations in unusual sequences – some of which Harel himself states “required some strenuous finger-twisting”.
Whilst Harel’s formidable Statecharts that describe the behavior of his Citizen digital watch obviously confirmed its status as a complex reactive system, they did not quite prepare me for the next generation of digital watches – or smart watches – which may in fact be closer to avionics systems in terms of their complexity.
If a digital watch with four buttons is a classic example of a reactive system, what is there to say about a smart watch that might incorporate a touchscreen, Bluetooth and WiFi radios, perhaps multiple cellular radios, light sensors, accelerometers, gyroscopes, NFC devices, barometric sensors, as well as a variety of other undisclosed sensors (but no buttons)? We are about to witness an explosion of health-related applications. Fitness applications that incorporate pulse rate measurements will become mundane. Recently the FDA approved an atrial fibrillation detection algorithm that gathers data from biometric sensors by way of a smart phone or wearable device. The cloud-based analytics engine can forward results directly to a cardiologist.
As digital watches have progressed from the original Texas Instruments LED models to the latest smart watches, I have been amazed at the successful evolution of these complex reactive systems. I remember teaching students the basics of reliability engineering. Complexity is the enemy of reliability. So how is it that today’s modern devices, which are hundreds of millions of times more complex than the earliest devices, are apparently more reliable than these earliest devices? The transistor count of Apple’s A7 processor exceeds 1 billion. The Android operating system consists of about 12 million lines of code. The full avionics functionality of the F-22 Raptor fighter aircraft is encapsulated in only 1.7 million lines of code!
Perhaps the most underappreciated engineering achievement is reflected in unprecedented manufacturing methods and technology that result in exceptional quality and incredibly low failure rates that contribute to the commercial success of these consumer devices.
Then there are the engineering techniques for managing complexity in design, implementation and testing. Harel’s Statecharts methodology is only one example of many techniques that allow for abstraction and isolation of functionality.
This combination of manufacturing technology and engineering complexity management is the driving force behind the constantly evolving digital watch that punctuates my life. It is difficult to imagine how forty years in the future, the new smart watch that we are so eagerly awaiting will be considered a quaint technology artifact – like the Texas Instruments LED digital watch. But based on the history and evolution of technology we can be certain that it will.
Obviously I need not be concerned that the quest for the ultimate digital watch is coming to an end. I seem destined to live in constant anticipation of the next cool digital watch. Just imagine the inane experiments I will be able to conduct with a next-generation avionics system strapped to my wrist! I am reminded of the opening paragraphs of Douglas Adams’ “Hitchhiker’s Guide to the Galaxy”: “Orbiting this at a distance of roughly ninety-two million miles is an utterly insignificant little blue-green planet whose ape-descended life forms are so amazingly primitive that they still think digital watches are a pretty neat idea.”
[1] Harel, David (1987) Statecharts: A Visual Formalism for Complex Systems, Science of Computer Programming 8 (1987) 231-274, North Holland
[2] Harel, David (2007) Statecharts in the making: a personal account, Proceedings of the third ACM SIGPLAN conference on history of programming languages, pages 5-1-5-43