well? are you going to get one?
Since I have an interest in ubiquitous sensors and computation many people assumed I'd get an Apple Watch the moment it came out. Something more interesting happened and I found myself with questions about what computation on the wrist is good for and the limits of what a wrist worn timepiece can do.
First I tipped my hat to Horace and starting thinking about the jobs to be done. Beyond cheap watches are camps of those who have an interest in different devices:
° a smartwatch that mostly tells time
° a wrist networked/computing device that happens to tell time
° a fine mechanical watch
None of these camps are right or wrong. In addition some people will use watches to signal social class or just enjoy the design and construction of a watch. Time keeping is central to some of these cases - particularly the mechanical watch - so it is useful to think of the jobs they were originally hired for - when innovations occurred and if innovation continues.
Navigation and event synchronization are important functions for accurate portable time pieces. Navigation is a fundamentally important driver of technology.1 Event synchronization became very important as the Industrial Revolution developed. There was a need to coordinate groups of people. Town clocks and home clocks that were synchronized to the town clock were important. The town clock would be set by the Sun's position - local noon could be determined and the clock could be adjusted regularly to good enough accuracy.
Railroads presented new problems. Every town had its own version of noon. Tables were produced showing the time offsets for clocks in other cities. Some stations kept clocks for each destination and hundreds of confusing charts were developed. Scheduling was a nightmare. Some crews would carry a half dozen clocks set for each major stop on a route. The clocks often lacked enough accuracy making matters even worse.
The solution was to standardize time. In 1840 Great Western Railway adopted London time. Within seven years most English railroads followed and finally a rail standards body recommended using Greenwich Mean Time. Local clocks throughout England followed, although many had two minute hands - one for local time and the other for GMT, and by about 1880 GMT was defined as the time throughout England. Other countries followed and time zones very roughly following divisions of longitude were established.
Pocket watches were reasonably accurate portable time pieces and the advent of standard time made them more practical beyond their original use on railroads. Some were beautifully built and expensive signaling social status in addition to giving the time of day. Wrist watches became accurate and durable enough to become important tools of military synchronization during WWI and mass production built an industry lowering prices to eventually make them a mass market item.
A decade before WWI a Einstein published one of his revolutionary results. He combined the experimental result that the speed of light was always constant with deep thinking about what the synchronization of events means and came to the conclusion that the measurement of time and that of space were not rigid. Moving clocks would run slowly and moving objects would shorten. Unless you could travel at a good fraction of the speed of light these effects were small. You would need a special clock that was much more accurate - more than a million times more accurate - than any known clock.
While I'm impressed by beautiful craftsmanship and design, I'm more taken with tools that allow revolutionary ideas to be discovered and verified. I see an Apple Watch or the best conventional watches and think about their accuracy compared to what it would take to verify special relatively. Such a watch would be impressive and more pure as a time keeping object. A regular watch does other jobs for some, but is only a time keeper for me.
Consider what it would take to verify the effect of special relatively by doing a thought experiment Einstein suggested. Place a clock in a moving vehicle and compare the time it shows to that of a clock that stays put with you. A nice trial would be to fly a clock around the world. Imagine an airplane going about 500 miles an hour on a 48 hour trip. A bit of calculation shows the airborne clock would show 47.99999999999867 hours had passed. You need a wicked good clock to see the difference.
It turns out atomic clocks are good enough. In October, 1971 the experiment was done by J.C. Hafele and R.E Keating. Two Hewlett Packard atomic clocks were flew in opposite directions around the world for about $7,600 in airline tickets for the crew and "Mr Clock". Analysis was complex as both Special and General Relativistic effects were involved.2 This was probably the least expensive new experiment involving General Relativity.
Portable watches won't give the required accuracy and why would you need it anyway? Watches can be synchronized GPS signals. They drift little enough between synchs that it is difficult to imagine a human level task they would fail at for lack of accuracy. But people have pushed the accuracy of standalone watches - atomic clock based watches and pocket watches have been built, but the atomic clock module they're based on isn't good enough to duplicate the 1971 flight around the world.3
The Apple Watch goes beyond time keeping allowing more convenient alerts and some health monitoring and encouragement. That isn't my use case - at least not yet. There will be many other sensors that it will be able to receive data from - a few hours thought gave me a list of a dozen that aren't crazy. That's where I'm hooked. If I had the time and inclination I'd be developing one or two now, but I rest assured many others have the same thought and some will be excellent developers.4
For me the Apple Watch represents future potential. When you see me with one consider it a signal that I've found something deeper that is beginning to get into the ubiquitous computation world.
1 I touched on it here. The subject is rich and one could write for hours. Fortunately many others have and the history of science and technology has rich stories.
2 Special Relativity for the relative motion of the clocks and General Relativity to adjust for gravitational redshift. For the eastward flight theory predicted 144ns gained from GR, 184ns lost from SR giving 40ns lost total (+/- 23) and 59ns lost +/-10 was measured.
The atomic clocks were stock HP-5061A Cesium beam atomic clocks made in the mid 60s. They weighed about 80 pounds each and required a separate power supply - semi-portable.
The paper is great fun.
J.C. Hafele and R. E. Keating, Science 177, 166 (1972)
3 DARPA funded development of a miniature atomic clock for GPS denied navigation. It was commercialized by Microsemi as the Quantum SA.45s. Very nifty and even inexpensive at about $1,500 for the module, it is useful in underwater navigation and some military applications where GPS signals are unreliable. As atomic clocks go it isn't very accurate - not enough for testing special relativity unless you are at orbital speeds.
There are many other questions. The world has standardized on an ITC time reference based on the Earth's rotation. Unfortunately this varies quite a bit, slowing down beyond tidal drag and leap seconds need to be added. The physics and geophysics involved is amazingly rich. We should probably move to just basing time on atomic clocks, but there is institutional pushback.
A quick note on time standards. Coordinated Universal Time (UTC) is a standard and not a time zone. All timing centers on Earth have agreed to synchronize their time scales with it. There are two components of UTC -
Universal Time (UT1) is solar time and is based on the Earth's rotation. It is adjusted several times a decade with leap seconds to adjust for unevenness in the Earth's rotation.
International Atomic Time (TAI) is a time scale that uses the output of about 200 precision atomic clocks around the world.
GMT is now a time zone.
Steve S. notes several issues with leap seconds:
- Some NTP/GPS servers implemented the leap second 5 months early.
- Google, Amazon (others?) are implementing smeared leap seconds
- The financial community are worried (this leap second occurs on a non-holiday weekday) and its overlord the Intercontinental Exchange have announced changes in the trading schedules for June 30 due to the leap second.
4 For the old Air Graffiti people who read this ... think iCane 2.0
Lentil Fruit Salad
° 2 cups red lentils
° 5 scallions chopped - just the white and white-green
° 1 red bell pepper sliced
° 1 medium red onion sliced
° 1 apple diced
° 3 oranges, 2 sliced and one juiced
° 1/2 cup olive oil
° white wine vinegar
° freshly ground black pepper
° cover the lentils with slightly salted water and bring to a boil over medium heat. Cook until tender - 45 minutes or so. Drain and allow to cool
° combine everything else in a large bowl and season with salt, pepper and white wine vinegar to taste
° mix in the lentils