Do people like, still blog?

Tag Archives: quantifiedself

When I started using a FitBit, I thought it was indispensable and really cool. Even though it only tracked steps and sometimes flights, I could have raw data on how I was moving! I didn’t have any real use for this information, but it appealed to the compulsive data nerd in me. I tried a bunch of alternatives to Fitbit but navigated back to the FitBit brand. Ostensibly, it was because the web and app interface was superior, but I think it was a whole bunch of little things that just made it easier to use.

This is a FitBit is good and FitBit is bad post. As with most things, the newness of the service wore off. I’ve been wearing a FitBit on my wrist for probably five years now. The data tracking is still there but I don’t really care for it anymore. The most useful feature is the slim profile that lets me see the time on my wrist. That’s not underselling it though, that feature really is useful.

I’ve gone through many FitBits. Started with the Force and then returned that for full credit as part of a product recall. Then I bought a first generation Flex and used it for awhile. The battery died (couldn’t hold a charge) and their support team replaced it with another Flex. Then the battery on that died the same way. Recently I bought a Charge for cheap on eBay (was new) and after three months, the band broke. Support replaced it once again, but because the product was so old, I received an Alta instead.

Their support is great but I wonder if there are just quality and design issues in their product line. When I owned the Flex, I probably bought and went through 20 different bands. They would just break after a few months of use. Luckily I bought them for cheap from China ($2-3 a pop) instead of the full retail price ($20+??). I can’t imagine the cost of ownership would be worth it otherwise. I also spent a bunch on chargers because different models had different connections.

The Alta I just received is the new hotness but I’m not sure I get enough benefit from it to deal with the accessories, charging (only 5 days of battery life) and the slow but eventual death of the hardware. Maybe I should just get a nice watch instead.

I was at a grocery store today and saw an ad for an electric toothbrush by Oral B. Now, most decent sized grocery stores in Toronto have a pharmacy section where they sell toiletries & etc, including a section for toothbrushes; but usually they don’t sell electronic toothbrushes (except for the < $10 ones you get from Colgate). So it was strange that they were advertising an electronic one.

Not only were they not selling the one they were advertising; it was an extra special toothbrush, which I’m guessing they’re trying to push for Christmas. They were advertising an electronic toothbrush that has built-in Bluetooth! I’ve never heard of one with BT and it seems a bit crazy for me that a manufacturer would tech that in. It piqued my interest, not because I wanted one, but because it’s so random.

I looked it up, and apparently it costs almost $250! You might get a cheaper one, or on sale but it’ll still set you back more than $100. Your typical high end, non-bluetooth-enabled power toothbrush seems to be around $50. That’s a big premium to pay for bluetooth.

Now, why would you want BT in a toothbrush? Obviously to measure how you brush your teeth. The toothbrush can measure how long you’re brushing and how much pressure you’re applying – and there’s a companion app to see how you’re trending over a period of time. I’m a big fan of quantified self, but I think this is going a bit too far. It’s not that hard to track this yourself (you can look at a clock to make sure you’re brushing for more than 2 minutes). I think it’s a gimmick and a cash grab hoping that people buy an extra expensive power toothbrush as a Christmas present. Don’t fall into this trap!

I have been anticipating the release of the 23andMe service in Canada for awhile now. If you haven’t heard of it, it’s a personal genome test service which takes some of your saliva and analyzes your DNA. I was interested in submitting my DNA for testing by for a long time it was a US-only service (possibly some ways to get around that). Then the FDA cracked down on the service so you could not get the full suite of their health analysis.

The rumors was that 23andMe would launch in Canada and the service here would not be limited by the FDA decision. Finally in October, they launched and I paid the $200 to do a test. 23andMe offers 2 basic services: 1) Genetic tests to see your risk factors for certain diseases (big ones are Alzheimer’s and Parkinson’s) and 2) Ancestry analysis. It turns out that they still don’t do as many genetic tests as they did prior to the FDA crackdown, and the set they do now (aside from the Big 2) are not that well known. I didn’t have a strong personal interest in the genetic tests, and had more interest in the ancestry analysis. But my main attraction in trying the service was curiosity in seeing what my DNA can reveal.

After purchase, the kit to collect my saliva came pretty fast (ordered on Friday, arrived on Monday). After spending 5 minutes spitting, I sent it off the next day. My saliva got sent to a location in Canada where it looked like it was bundled with some other kits and sent to their processing facility in the US. That took about a week (it seems like it had to go through custom clearance due to biological material?). Then another week for them to do the analysis. All told, it took about 2.5 weeks to get my results.

The results were a bit underwhelming. The don’t test as many things as I thought they would in terms of genetic characteristics (i.e., whether you are lactose tolerant or not) and the number of genetic risk factors is also small. There’s always promise that they will add more risk factors and your DNA analysis will automatically be applied to those. The ancestry result was also not too interesting, although that might be because the result lined up with what I thought my genetics would be. There’s a feature to find DNA relatives, and 23andMe found a couple – but they were all beyond 3rd cousins so we are not really related (of course this feature depends on the number of people using 23andMe).

For the price, I don’t think you get sufficient value out of it (especially if there are no surprises). However, I am interested in progressing the field and my understanding of my DNA so the $200 to become a member of this service is my investment in that.

Although the Hackathon which I entered A Healthier Commute has completed, I’ve been continuing to work on the product. It’s interesting to me because I want to do some research into whether I can track commutes using just a phone. Well, certainly you can do it, which is what apps like Waze or Uber (in reverse) do. You could even use something similar like Strava or MyTracks to do it. But I think what’s different is that all of those require active user intervention.

I have a FitBit Force and it has this neat feature that does sleep tracking. It collects data, processes it and spits out all sorts of neat conclusions like your sleep efficiency. You can see how long it took you to fall asleep, whether your sleep was restless and how many times you were awake. The thing I hate the most about it is that you need to remember to tell it start tracking when you go to sleep. Not only that, when you wake up, you have to remember to tell it to stop tracking. It’s very easy to forget one or the other.

I think it’s the same with commutes and/or driving. You might come up with a clever marketing strategy and get people to track their commutes for a week; but after that people tend to forget or get tired of doing things. And it’s the aggregate data that holds the most value, so it’s important that people keep tracking for longer than a week.

So with A Healthier Commute, I’m trying to make all this tracking automatic and passive. No buttons. It’ll just be smart enough to display your commutes. That’s challenging, difficult and interesting.

A few weeks ago, I happened on some news that said that there was a voluntary recall on the Fitbit Force. It turns out that retailers had stopped selling the Force as well too. I have a Force so I was curious why this was happening, and it was quite easy to find out. Basically, 1.7% of people reported skin irritation (although the number might be higher as not everyone reports problems). I didn’t get the problem and was debating whether to return it.

In the end I figure I will return it (still waiting for my return kit). Although it’s an interesting toy, and it’s interesting to know my activity and sleep patterns; I am not enamoured enough with it that I desperately need it. It’s not every day that you can try out a product and then return it for a full refund, so that in itself is a solid reason (and I can put the money towards the next model).

Wearing a Fitbit also bothers me a bit because now that it is on my wrist, I can’t wear a watch! In fact I was dreaming of a better FitBit device which would latch on the frame of your glasses. It would be unobtrusive and would go where ever you go (I guess it wouldn’t track sleep that well).

I think the person who will miss my Force the most is Apollo, he loves being able to turn the OLED on and watching it animate off.

This past weekend I participated in a hackathon organized by the Government of Canada surrounding its Open Data initiative. I am enthusiastic about Open Data because it provides correct, free and consistent data that I can build upon, which is unlike most other apps I build.

I had been thinking about what sort of app I would make for awhile. I was quite positive that I wanted to do something related to Quantified Self because that is something I’m interested in right now. When the theme of solving problems and improving productivity was announced; I decided that I wanted to tackle the problem of congestion on our roadways.

I think the major contributor to congestion is just people commuting to work. Especially people by themselves in a car. My idea was to provide personalized feedback and show how much money, time, and environmental impact you cause by doing that. The Open Data was critical because I was able to get data points for specific cars that users would drive.

The UI was fun to make up. I think I discovered a clever little hack which made my life a lot easier. First, I asked the user to specify the start of their trip (probably home) and the destination (probably work). Then I made a call to Google Maps to pull up directions between those two places. It may not be the exact path they take to work, but it would be close.

From there I calculate the impact. I thought the most interesting one was the CO2 emissions. Usually you just get a number, but hearing a number is meaningless. I translated that number into a measure of garbage trucks. It alludes to the fact that you are generating garbage, and makes the result easier to comprehend.

I’ve published A Healthier Commute on Google Play so you should check it out if you haven’t yet!

I’ve been using my Fitbit for about half a month now. I’ve been pretty good at using it constantly, just taking it off to shower (it’s supposedly waterproof, but there is a tiny hole for the altimeter which can possibly let water in if the water pressure is too high). Here are some of my observations:

  • The step tracking is not accurate at all. I even changed to setting so that it’s on my “dominant” hand even though I’m wearing it on my non-dominant hand. It records steps even though I am just working at my desk. The Fitbit marketing material even says that it’s not the actual number that counts, but the trend. I think this is why some competitors (like Nike) don’t show the actual steps.
  • The initial 10,000 step per day goal is really tough. I can only make it on the weekends when I walk around a mall for a few hours or have to run a lot of errands
  • The floor climbing seems to be pretty accurate, assuming you know two things: 1) It tracks entire flights and not individual stairs), and 2) It only tracks flights that you climb (not decline).
  • Living in a house, the initial 10 flight goal is way too easy. I increased it to 25 flights, and I can still surpass the goal on most days
  • The sleep tracker is pretty neat. Initially, I thought it was pretty accurate, but I’m not entirely sure that it is true. For example, when I’m trying to sleep, I don’t move my hand around so I’m not sure how it still knows that I’m restless/awake.
  • I think the sleep tracker and the stats that come out of that (how long you sleep on average, how many times you wake up etc) are the most interesting data that I get out of the Fitbit

For the most part, the Force is still interesting to use. I’m not sure whether I’ll continue to, or remember to, turn on/off the sleep tracker every night after awhile though.

I am using a Fitbit Force nowadays as part of a focus on quantified self for work, but of course the data that it records is also interesting to me. It is a strange device if you think about it, because it’s really a glorified pedometer. If you truly wanted a pedometer, you could buy one for a couple of bucks at the dollar store. The Force will set you back $130.

From a tech POV, it’s novel because there aren’t many devices that are similar (but there are other players in this market too). It’s a piece of wearable tech that is beneficial and not too expensive. It packs a 7-day battery, OLED screen, vibration mechanism, pedometer, altimeter, NFC and bluetooth radio into the rubber wristband. I decided to get the Force over the previous model (Flex) because the Force had a clock display, and in general, a more detailed display (you can see actual counts instead of just a relative-to-goal indicator)

However, most of the tech is auxiliary, and the primary function is that of a pedometer. The hardware itself is not necessarily worth the 30x or 50x price difference between a dollar store version, but there is also substantial software around it. The Force syncs wirelessly with your device and uploads your data to your online dashboard. You can set goals, see your progress, and track your behaviour across days, weeks and months. I think that’s interesting and perhaps worth the premium.