Category Archives: gadgets

2 Useful Insights I’ve Gained from Personal Analytics Data: Sleep and Productivity

In writing about personal analytics and data collection, one question I get more frequently than most is: what do you get out of it? Today I thought I’d share 4 insights I’ve gained into my own behavior from scrutinizing the data that I collect.

For those who haven’t been following along, I am fascinated by what data about our everyday lives can tell us about our behaviors. The data is often referred to as “personal analytics” and the movement behind this kind of data collection and analysis is called the “quantified self” movement. I collect data in four major areas:

Areas of Tracking

I collect data in other areas, too, but the key point about these four areas is that the process is entirely automated. I just go about my day, and this data is collected without any intervention or action on my part. I’ve already written extensively about my walking and writing insights so today I’ll focus on what I’ve learned about my behavior when it comes to sleeping and overall productivity.

1. Restless nights and sleep efficiency

You know those nights where you feel like you are tossing and turning all night long, getting very little sleep? Turns out, I do sleep on those nights, at least according to my FitBit, but my “sleep efficiency” is down below 90%. Here is a one recent example:

Sleep Efficiency

I’ve been capturing this type of data for almost two years now and I’ve learned a few useful things about my sleep habits by looking closely at the data.

  1. When my sleep efficiency is >= 95%, it feels like a restful night’s sleep. This is true for me almost independent of the number of hours I actually sleep. If I only get 5 hours of sleep, but my sleep efficiency is, say, 97%, I still wake up feeling like I had a good night’s sleep.
  2. When my sleep efficiency is between 90-95%, it’s a pretty good night, but the number of hours is more of a factor. If I get, say 7 hours of sleep with a sleep efficiency of 92%, I feel pretty good in the morning. On the other hand, if I get 5-1/2 hours of sleep with a 92% efficiency, then I don’t feel nearly as well-rested. According to the data, the time threshold is around 6 hours.
  3. When my sleep efficiency is less than 90%, I feel like I had a restless night’s sleep, regardless of hours actually slept.

I’ve been able to take this data and put together a chart of my sleep quality, based on two variables, sleep efficiency, and hours of actual sleep (vs. hours in bed).

Sleep Quality

I should not that I do not track how I feel each morning when I wake up. But on mornings when I felt particularly good or poor, I’ve checked it against the data from my FitBit and it is fairly consistent. For me, therefore, the above chart is a good representation of the quality of my sleep based on the two inputs.

How does this help?

Continue reading 2 Useful Insights I’ve Gained from Personal Analytics Data: Sleep and Productivity

Our New Nikon D5100 D-SLR Camera

This year, our Christmas present to each other was to get a new digital camera. In the past we’ve always had cameras like the Canon PowerShot, perfectly good cameras, but I’ve been wanting an SLR camera for a while now. We didn’t need anything super-fancy, but I wanted something that at least had good recommendations. On his blog, John Scalzi mentioned using a Nikon D5100 in a post on his photo tools. Starting there, I looked into the camera, and it seemed to have everything we could want.

And there was one other piece of serendipity. We ordered our camera on Cyber Monday, and in doing so managed to get the camera for nearly 50% off its regular price. We ordered the camera as part of a kit that included 2 lenses: an 18-55mm and a 55-200mm lens. It was an incredible deal!

Well, since today is Christmas day, we finally got to play with the new toy. I read through the manual–something that I almost always do with new gadgets–and then started taking pictures. Nothing fancy, just to get a feel for the features and how the camera works. There are a ton of features built into the camera. Here are two of the pictures I took with the new camera today. The first is just a regular photo; the second is using some of the camera’s special effects.

Christmas Tree

And here is the same scene with a built-in special effect that I think is really cool.

Special Effect

I have no illusions that I am a photographer of any particular skill. I leave that to people like Paul Weimer, Andy Romine, or the aforementioned John Scalzi. But it is nice to have a camera that can help make my pictures look better.


Initial Thoughts on the Automatic Link Smart Driving Assistant

A while back, a friend of mine, knowing my penchant for personal analytics, call my attention to a new device being manufactured by Automatic called the Automatic Link. This device captures data about your driving the way a FitBit device captures data about your activity. At the time, it was not available, but it eventually became available, and a few weeks ago, I got my hands on one. I had it just in time for our trip down to Florida, which is where I really put it to use.

The device is designed to plug into the data port on your car that mechanics use to determine problems when the “check engine” light is on. Automatic says that it works with most cars manufactured since 1996. Once plugged in, the device syncs up via BlueTooth 4 to your iPhone and provides all kinds of data about your driving. It captures this data in real-time by communicating with your car’s computer1.

Automatic Link in my Kia
Automatic Link plugged into our Kia

There is an initial setup that syncs the device to your phone. After that, no action is required to track your driving. If the app is running on your iPhone, it automatically detects when the ignition starts and stops, and tracks everything for you. The primary goal of the Automatic Link seems to be saving your money by making you a more fuel efficient driver. Because of this, the device tracks three things that makes you less fuel efficient:

  1. Fast accelerations
  2. Hard braking
  3. Speeds over 70 MPH

You can set an audible alert for each of these so that the device notifies you when any of these things occur in real time. Since I started using the device, I have been dinged for all three, and I can attest that, at least in my car, the alerts are accurate.

The device also tracks your route and mileage and can compute how much your are spending in gasoline costs by being aware of local gas prices. Here is what the first leg of our drive down to Florida looks like in the Automatic Link app on my iPhone:

1st Leg

The device captured our departure time (9:03 am) and where we were departing from (“Home”). It recorded 121.2 miles before engine shutdown a little over 2 hours later. You can see from the section I circled about that I had 1 hard brake, 1 hard acceleration, and spent 37 minutes over 70 MPH. This latter was on I-95 where the speed limit is 70 and I had the cruise control set to 72 MPH. It calculated the cost of gas for that leg at $14.47.

You can click on a trip to get additional details, including a map of the route taken:

1st Leg Map

Here you can see the actual route. I annotated the screen capture with arrows that show exactly where the hard acceleration happened, and where I was driving over 70 MPH. Another cool feature is that you can see your fuel efficiency (circled above). In the case of this first leg, our Kia Sorento got 29 mpg.

Our longest single leg of driving on our way down to Florida looked liked this:

Continue reading Initial Thoughts on the Automatic Link Smart Driving Assistant

  1. Indeed, it reminds me of that scene in Empire Strikes Back when Han Solo asks C3PO to talk to the computer of the Millennium Falcoln in order to find out what’s wrong with the hyperdrive.

New Toy for a Data Hound

I got a new toy today, the perfect thing for a data hound like me. It is the Automatic device, which is probably best described as a “FitBit for cars” type of device. The Automatic website describes the device as, “Your smart driving assistant.” And it comes just in time for our drive down to Florida next weekend.


The device plugs into the standard OBD-II data port on your car (it is compatible with most cars made after 1996) and syncs via BlueTooth 4.0 with the Automatic app on your iPhone. It does four main things:

  1. Provides real-time driving feedback on things like rough braking, speeding, and rapid acceleration, all of which contribute to lower fuel efficiency. It calculate a driving score for you and tells you how much money you save on gas based on how you drive.
  2. It creates trip timelines for you with all kinds of information about your driving habits, where, and how much you drive. It can detect when the car should be filled up and knows about local gas prices so that you can save money at the pump.
  3. It remembers where you park your car.
  4. It knows all of the codes for the “Check Engine” light, should it come on.

The device can also contact emergency services if the car is in a serious crash. (It has a built-in accelerometer.)

Of course, for me, this is just another source of automatically collected data for personal analytics that I can add to my FitBit activity data, writing data, and other data that I collect. I’ll see how it works on our vacation and you can bet I’ll be posting more about it.

Does a FitBit Flex Accurately Track Your Steps if You Walk with Your Hands in Your Pockets?

It was bitterly cold here the last few days, making my lunchtime walks a little more challenging. I normally walk 3 times around the large block where my office is located, which amounts to about 3 miles, or roughly 6,000 steps on my FitBit Flex. With the weather as cold as it was, I was a little concerned about how my FitBit Flex would behave if I shoved my hands in my pockets. Earlier in the spring, I discovered how pushing a stroller skews the number of steps you take. How would shoving my hands in my pockets to keep them warm affect the way the Flex counts my steps?

In the interest of science and discovery, I decided to perform an experiment yesterday. I would walk each of my 3 laps at lunch in a slightly different manner:

  1. First, I’d walk the entire lap without putting my hands in my pocket.
  2. Next, I’d walk the second lap with my hands in my pocket the entire time.
  3. Finally, I’d walk the third lap as I always do, with a kind of hybrid mix of the first two.

I used the activity-tracking feature of the FitBit Flex to capture the data from each lap independently. When the experiment was concluded, here is how things looks:

Winter Walking
Click to enlarge

Let me examine each lap independently.

The first lap – hands out of pockets

On the first lap, I walked with my hands out of my pockets for the entire lap. I walked just about 1,800 steps in 14 minutes. The first thing you’ll note is that this is two minutes faster than the other two laps, each of which took me 16 minutes. I think there is a very good explanation for this. With the windchill, the temperatures were below freezing. I was walking ungloved with my hands outside my pockets. It was cold and I therefore walked faster.

The second lap – hands in pockets

On the second lap, I walked with my hands in my pockets the entire lap. I walked just over 1,900 steps in 16 minutes. I have already explained why this lap was longer by 2 minutes (with my hands in my pockets, I wasn’t as cold and my pace was slower). The same walk took about 100 more steps with my hands in my pockets. 100 steps is roughly 5% of the total, which one could argue is within a margin of error of +/- 5%. But I think there are 2 other possible explanations:

  1. Putting my hands in my pockets alters my stride, making it a little shorter. Over the course of 2,000 steps, this adds about 100 steps to make up the distance.
  2. I didn’t walk the exact same course. Each time around is a little different, weaving this way or that to avoid obstacles, detour around people, etc.

Each of these I think easily explains an extra hundred steps. The good news, for me at least, is that the step count did not drop with my hands in my pockets, the way it does when I push a stroller.

The third lap – hybrid

The third lap is closer to how I normally walk in cold weather, shoving my hands in my pockets for a little while, and then taking them out for a little while. As you can see from the data, it is almost identical to the second lap. This tells me 2 things:

  1. The first lap was an outlier, the numbers altered more because I was cold and wanted to rush through the lap.
  2. The 3rd lap differed from the 2nd by only 36 steps. This tells me that whether or not my hands are in my pocket, my FitBit Flex is recording my steps the same.


It seems, based on my simple experiment, that walking with your hands in your pockets in cold weather does not affect how the FitBit Flex records your steps. You don’t have to worry about “losing” steps in the way you lose them when pushing a stroller. As winter approaches and cold weather becomes more common, you can rest assured knowing that the Flex is capturing your steps accurately, whether you walk with your hands in your pockets, or outside them.

5 Terabytes of Local Storage at Home

Recently, the 1 TB external disk attached to my iMac exceeded 90% of its capacity. I’ve had this external drive for almost 5 years and it has taken me that long to fill it up. The drive contains all of our media files (iTunes music, movies, TV shows), locally stored photos (iPhoto), and software installation binaries. The original external drive was a 1 TB WD My Book. It has been very reliable for the entire time I’ve had it, so earlier this week, I ordered another one and it arrived today.

The new external disk has a 3 TB capacity. I moved all of the media files (music, movies, TV shows) to the new drive, and left the photos and installation binaries on the old one. This should give us plenty of local disk capacity into the foreseeable future. Indeed, if you count the 1 TB drive that came with my iMac, we now have a 5 terabytes of local storage capacity. I can still remember 20 years ago when 40 megabytes seemed like an enormous hard disk for your PC.

I’ve configured to the new external disk to be backed up by CrashPlan (not everything, just those things that are not synced to the cloud via iTunes Match) and I imagine it will take a day or two for those backups to run. In the meantime, it feels good to have the new space.

5 Tips for Getting the Most out of a FitBit Flex

I have had my FitBit Flex for something like five months now, long enough to grow comfortable with its benefits and quirks and so I thought I’d put together a list of some tips that can help people who have one (or are thinking about getting one) get the most out of it.

1. Wear it and forget it

Perhaps the best thing about the FitBit Flex is that you wear it on your wrist. No need to clip it onto your clothing the way you had to with a FitBit Ultra, for instance. And since the FitBit Flex is waterproof, you can shower with it, meaning there is little need to take it off. This means you can wear it and forget it. You won’t find that you’ve left the house for the day only to leave your activity tracker on the dresser.

2. But set a recurring reminder to charge it

That said, I have found that I need to charge my FitBit Flex every 5-6 days. The battery has only run completely down once. The way I avoid the battery running completely down is by proactively reminding myself to charge the battery.

I use the reminder feature in Google Calendar and have created a recurring reminder to send a text message to my phone every 5 days, reminding me to charge my Flex.

When I get the reminder, I typically charge the Flex in the evening when I am doing my writing and getting the kids ready for bed. It doesn’t take more than 2 hours or so to get a full charge and by the time I go to bed, the device is once again fully charged.

3. Track specific activities

You can (with the latest firmware update) track individual activities with your Flex, which is useful if you want to know some detail about that activity. You do this by putting your Flex into Activity/Sleep mode (tapping it rapidly until you get the two flashing lights). When your activity is finished, you tap it rapidly again until you see 5 blinking lights.

What this does is sets a “start” and “end” marker. Your FitBit is constantly recording your activity and these markers allow you to look at a specific section of that activity to see the details. When you log into the FitBit dashboard, you will see those activities listed in the day on which you captured them. I’ve used this to capture the distance of my walk from my house to my office, for instance. Or how long a particular hike took. Here is what an activity record looks like for my lunchtime walk, as seen from the FitBit website dashboard:

Lunchtime Walk

Continue reading 5 Tips for Getting the Most out of a FitBit Flex

Here’s A Picture of the Line At the Apple Store Outside My Office

There has been a long line of people waiting to get new iPhone 5{c|s} devices from the Apple Store in Pentagon City today. Here is a picture of the line, taken from my office which sits in a building above the food court of the mall, looking into the rafters. Enjoy it and be glad you are not standing in it yourself.

Apple Line

My Fujitsu ScanSnap s1300i is Still Awesome One Year Later

I recently realized that I have been using my awesome Fujitsu ScanSnap s1300i scanner for a little over a year now. One year later, it is still awesome and one of the most important tools I have to keep me paperless.

Given that I’ve been using the scanner now for over a year, here’s a few thoughts on my experience with the scanner and how things have changes in the year that I have been using it.

  1. The s1300i is still the primary way I go paperless, converting any paper I get into digital form by running it through the scanner and getting rid of the paper.
  2. In the year that I’ve been using the scanner, I cannot recall a single occasion upon which I have had a problem. The scanner rarely jams; its paper feeder is excellent. And on the rare occasions it has jammed, it has almost always been user error.
  3. I have learned to speed up the scanning process by turning off the features in the scanner software that create searchable PDFs. This slows down the scanning, and once the document is scanned to Evernote, the resulting PDF will become searchable, thanks to a similar feature that runs on Evernote’s servers. This allows me to scan paper in quickly, but still get searchable documents out of them.
  4. Part of what makes the scanner so successful is that it is incredibly simple to get documents into Evernote. I set the paper in the feeder, push the scanner button, the document is scanned and it automatically appears in Evernote. Done.
  5. When I started using the scanner, I had a bin on my desk in which I’d collect the paper I needed to scan. A year later, I no longer have the bin because I no longer get that much paper. Today, when I have paper I need to scan, if I don’t scan it right away, I simply set the paper on top of my s1300i, and it is there waiting for me the next time I am ready to scan.

I think we often take for granted tools that just work without any problems. Televisions and refrigerators are two such appliances that I’ve had little occasion to complain about. And my ScanSnap s1300i is another one to add to this list. It just works, no problems, and has done so for over a year.

I wanted to try to figure out just how many pages I’ve scanned in that year, but I ran out of time. Extrapolating from some data I pulled from Evernote, however, I think I’ve scanned on the order of 500 documents, probably totalling somewhere between 1,000 and 1,500 pages. The frequency of scanning has gone down considerably because I get less and less paper, but that still averages to between 3-5 pages per day.

Kudos to the Fujitsu people for making such a reliable, useful scanner. It has been a big part of why I have been able to go paperless.

Three More Tips for FitBit Flex Users

I have been using my FitBit Flex since late April now, and it has been a mostly great experience. Unlike my old FitBit Ultra (which I ultimately lost) I don’t have to worry about losing it because I wear it on my wrist. I also don’t worry about forgetting it in the morning because I sleep with it on each night in order to track my sleep. The data is provides is great, and particularly useful for someone like me, who enjoys playing around with the data.

But I did say I was mostly happy. There have been three problems, both of which have been corrected, but which I wanted to mention here in case anyone else has experienced these problems.

1. Sudden drop in step count, despite the same amount of walking

Not long after I got my Flex, I upped my daily goal to 15,000 steps/day. At first, I had a little difficulty meeting this goal, but once I began spending my entire lunch hour walking, it became routine not just to meet, but to exceed this goal on a regular basis. A typical week of walking looked like this for me:

Steps - June 2
FitBit Flex steps record for the week of June 2, 2013

A few weeks later, beginning around June 23, I noticed a sizable drop off in the step count for the week:

Steps - June 23
FitBit Flex steps recorded for the week of June 23, 2013

Set aside the two days at the end of the week when we were driving up to Maine and I was in the car for most of the day, doing very little walking. What is odd about this week is that my step counts are down across the board, never even reaching 15,000 steps.

But the strangest thing of all is that I was taking the same 3 walks I do every day. I was not walking any more, but I was not walking any less. So why the lower counts?

Take a look at the data for Tuesday, June 5:

Steps on June 4

You can clearly see the steps for my three walks. You can further see that each time, they reach nearly 600 steps for the time interval shown, which I think is 5 minutes. Now take a look at the data for exactly 3 weeks later, the week in which I noticed lower step counts:

Steps on June 25

You can still see all three of my walks, but you’ll note that they barely reach 500 steps/interval, despite the fact that I tend to walk at the same pace and the same distance each time. For some reason, my Flex was under-reporting steps.

Of course, it could have been over-reporting for all of the other weeks, but since this week was the exception, I decided that it was under-reporting.

I tried a number of things, none of which seemed to make a difference, including resetting the device. We’ll come back to this shortly. First…

Continue reading Three More Tips for FitBit Flex Users

Does Pushing a Stroller Affect FitBit Flex Data Capture?

We spent our Memorial Day afternoon hanging out with friends. Kelly took the kids to the local park to meet our friends while I walked to the nearby barber to get a long-overdue haircut. From the barber I walked to the park to meet-up with everyone. After a while, we decided to get something to eat and walked to a nearby Mexican restaurant. Lots of margaritas were consumed and great fun was had by all.

But yesterday evening, when I looked at my FitBit data, I noticed something strange. Although the walk to the Mexican Restaurant was about 1-1/2 miles each way, it seemed that my FitBit Flex had not captured those steps. It had captured the steps when I’d done roughly the same walk a week ago. Here is a profile of my steps for last Sunday:

FitBit Flex - No Stroller

The red arrows indicate the to- and from- of my walk last Sunday. You can see that I started my walk just before 10am. I walked about 2 miles, took a break (that small red line in between the tall green lines) and then walked back. Now, take a look at roughly the same walk yesterday:

FitBit Flex with Stroller

I’ve indicated where I walked to get my haircut and then to the part. The two arrows with question marks are where we all walked to the restaurant. Notice there is almost no steps recorded during nearly 3 miles of walking. But then, at around 6pm, my walk to the nearby grocery store is captured.

I was a little annoyed to lose those steps, and it occurred to me that my steps over the weekend, despite doing a lot of walking, seemed unnaturally low. So I started thinking if there was anything in common during those times when the steps were not captured. And there was!

During the time we walked to- and from- the restaurant yesterday, I was pushing a stroller. I was also pushing a stroller quite a bit over the weekend. But I was not pushing the stroller when I walked to get my haircut or when I walked to the grocery store. It seems to me, therefore, that the FitBit Flex has some problems capturing steps when the arm on which you wear it is fairly steady–as when you are gripping the handle of a stroller or shopping cart.

I planned to contact FitBit about this but checked their FAQs first, and found this tidbit in their accuracy FAQ:

Will Flex count steps when I’m pushing a stroller or a shopping cart?

Flex will count your steps when you are pushing a stroller or a shopping cart as we do want to give you credit for this activity. That said, because your hands are not moving, your step count may be a bit lower during this activity.

Although it is a little frustrating to lose all of those steps (I’d estimate that over the weekend, in excess of 8,000 steps are missing from my data because I was pushing a stroller), it was interesting to make and confirm this discovery entirely from the data and independent of the FAQ. The fix is pretty easy–push the stroller with one hand. And most importantly, I know that while FitBit reported my activity was lower, I was still getting the exercise I need.

My FitBit Flex and Sleep Tracking

One question I’ve been asked on several occasions is just how accurate the sleep tracer in the FitBit devices are. I used a FitBit Ultra for over a year and have been using the FitBit Flex for a couple of weeks now. I’ve used the sleep tracker in both, but unlike the Ultra, which required an awkward armband, the FitBit Flex is always on so it is easy to use the sleep tracker.

My experience has been as follows:

  • On nights when I feel like I’ve slept pretty well, the FitBit Flex shows that by showing that I didn’t wake up very many times, or wasn’t very restless.
  • On nights when I feel like I haven’t slept well, the FitBit shows more restlessness.

It is hard to put this to the test because I tend to sleep in waves of good or bad. That is, when I am sleeping well, I’ll sleep well for weeks on end. Then, when I’m not sleeping well, that will go on for weeks as well.

Last night, however, seemed like a very poor night’s sleep from my perspective. I was tired, but restless. I felt like I was looking at the clock every few minutes and really didn’t settle down into a deep sleep. I figured that last night would be a good test to compare my perception of my sleep to what the FitBit Flex recorded. Here is what the FitBit Flex had to say:

Sleep Chart

The red lines show when I was awake. The light blue lines show when I was “restless.” Looking at this, I think it reflects very well my restlessness throughout the night.The first red line, right around 11am, was when the Little Man came into our room to tell me he needed to use the bathroom. (This is a little difficult operation for him to perform at night by himself with his cast on.) The second red line, right around midnight, was when I awoke after an unusual dream1 and jotted a note about the dream so I wouldn’t forget it.

But look at all of those light blue lines, scattered pretty evenly throughout the night. It seemed to me that I was tossing and turning for most of the night, and that is exactly what the FitBit captured. Still, a more quantitative number would be helpful. When you drill into the details for the night’s sleep you see something like this:

Sleep Details

Here, you can see that I fell asleep in 4 minutes, was awakened 17 times (some of those times count as “restlessness” as opposed to being awake, the difference being how much movement there is each time. I was in bed for 7 hours and 54 minutes and asleep for 7 hour and 16 minutes. This makes my “sleep efficiency” for last night 93%/

I have sleep efficiency data for other nights, and it seems to me that when my sleep efficiency is above 96%, I feel like I’ve slept well. When it is between 94-96%, I slept “okay.” But when it falls below 93%, I feel like I’ve had a rough night.

I think this backs up my feeling that the FitBit does a reasonably good job of capturing the quality of your night’s sleep. The resulting data matches my perceptions very well. So when I am asked how well the FitBit tracks the quality of my sleep, I guess my answer is the same as it has always been:

It does a very good job, as far as I can tell.

  1. About which I’ll have more to say in a subsequent post.