Hi, I'm Rob Lillack, proud father of three and a software architect from Leipzig. I like motorbikes. Oh, and there's some cool software I created. My twitter username is @roblillack.
Posts tagged “running”

The Million Meter Challenge

On January 1st 2014, I decided to take another stab at running. Since then, I have been running every single day. Here’s how I set the right goal for myself.

I’d been running on and off for years now, sometimes even taking part in smaller races without any real achievement. I had ran my first 10K in 2011 without any preparation. It was a near-death experience with an average heartbeat of 194 during the one hour and ten minutes long run. My personal best was from November 2012, an unbelievably fast time of 55 minutes I reached after three months of hard training. At the yearly business race event “Firmenlauf” in Leipzig, I’d usually be the slowest of the company.


The thing is: I can easily motivate myself to try to achieve any kind of comprehensible goal. But getting fit, slimming down that belly—those kinds of fuzzy objectives had been completely out of reach for me.

Being the leader and manager of 40-something software engineers I knew motivation is an ever-intristic thing and cannot be imposed upon somebody but merely be activated at best. I knew that this activation can only happen using SMART goals—but somehow, I never thought of applying those criteria to personal projects, let alone, my body.

Trying to find some goal that would ‘trick’ me into getting fit, I came up with something very easy to grasp: to run 1,000,000 meters during the year 2014. Simple as that.

Let’s check the SMART criteria:

  • The challenge is SPECIFIC: There’s no messing around with copious goals that might be cool to achieve but aren’t really easy to explain to your weaker self; like having a sixpack again, or feeling better, or going to gym more often. It’s just running. A specific number of meters. Perfect.
  • It is MEASURABLE: Again, one million meters. Not “more often” or something. It doesn’t matter how, if it is a million in the end.
  • It is ACHIEVABLE: At least, that’s what I thought at that time. Anybody can run 2.74 km per day, right?
  • It is RELEVANT: Well, here’s the trick part: Running 1000 km actually is not relevant at all. But in the context of trying to live a fitter, healthier life, running often enough to reach that goal is pretty much the most relevant thing I can think of: If everything else fails, but I’d still be running, I wouldn’t need to worry about my health.
  • It is TIME-BOUND: Even today I still don’t now, what will happen in 2015. When thinking about the challenge, I tried not to force me to be running for the rest of my life—I just hoped, I’d be addicted by the end of the year.

After a few days of running, I noticed a fear coming up: What if I was getting ill and needed to recover a few days? And what if I didn’t really feel well enough to start running again for a long time? I’d run up ‘meter debts’ and would need to try and run longer distances to get debt-free again. This in turn would increase my risk of injury or getting sick …

The answer was clear: I would just run every single day. Regardless of the country I happen to be in that day, or if my wife would be delivering our baby. It doesn’t matter how I’d feel, as long as I was able to run, I’d just give my best and tweet the results. This was my secret strategy to fight my weaker self and my tactic to reach the goal.


Turns out, all this self trickery worked pretty well: Today, 100 days early, I crossed the finish line. Until now, I burned 71,717 calories during more than 108 hours spent running 1,003 kilometers—and I can’t think about stopping anytime soon.

I’ll write another post about how I measure my progress and my ultimate goal—fitness—to make sure the challenge is, in fact, relevant.

For the time being, make sure to check out my running log on twitter (I also have a real account) or discuss this post on reddit.

Testing smartphone GPS accuracy

We all know the iPhone 4s is one of the fastest smartphones available when it comes to getting a GPS fix. I have seen Android devices (OG Droid, I’m looking at you!) taking THREE whole days to find out it was just brought to the middle east, while the 4s answered the turning-off of flight mode outright with “you’re still on this runway, head into that direction to get to the cabs” while the flight attendants where still fiddling with the plane’s door.

But how do current smartphones fare when it comes to accuracy of the determined location? Can we really throw our standalone GPS devices away now?


Today I recorded a small walk with my family using a Garmin Forerunner 210 GPS watch (this is actually a fitness device including a heart rate monitor strap) and two smartphones: the Apple iPhone 4s and a BlackBerry Bold 9900.

On both phones, Endomondo SportsTracker was used to record the walk. It took us about one hour to get to Leipzig’s Zoo, grab a bratwurst, take a peek at the monkeys, and rush home again because it started to rain. Our average pace was somewhere around half an hour per kilometer—with two kids bringing doll’s prams this is just as fast as it gets.


Shockingly, it was not the iPhone that made the race here. While, of course, the black slab was the fastest to find enough satellites (with the BlackBerry coming in 2nd place), the quality of the recorded track is the worst of all three devices.

Even the GPS watch sometimes did not exactly hit the nail, but still, a smartphone cannot beat the accuracy of a standalone device right now. Whats worse with the iPhone: The timing seems to be a few seconds off and there are some more-or-less heavy misfires amongst the recorded track points.

The difference can be clearly seen on this video comparison:

Being a bit more “jumpy” when recording GPS tracks may not sound that bad at first, but in this case, the iPhone recorded a track that is nearly 25% longer(!) than what the Garmin delivered. The BlackBerry’s, on the other hand, turned out to be 5% shorter that the watch’s.

Of course, these are all not scientific measures, I don’t know for sure how long the path really is, we where probably moving ten times as slow as the average jogger, the sky was rather clouded, and the distance we walked was also nothing really fancy—but still, this is something to watch out for if you want to count on your iPhone for your running workout.


The lower recording detail is probably a software issue and other apps may be available, which record tracks with a higher frequency, but in the current test both phones only recorded a fifth of the number of track points, compared to the Garmin: That’s around two per minute for Endomondo, and one every five seconds for the GPS watch.

The smartphones did not record any height information (or Endomondo did not export this) so I did not compare anything like this. Also, I’m very interested in how the current Lumia devices and the N9 from Nokia fare—but I don’t have one at hand right now.

Conclusion: If you have a standalone GPS unit, continue using it! If not, then your BlackBerry works fairly good, too. If you want to use an iPhone to measure your jogging trail, better make sure you’re able to edit out any jumps afterwards, if needed.