Back to work

Well yesterday sure didn’t feel like much of a day off.  The weather was glorious but I spent only limited time outdoors enjoying it when I changed the oil in the Prius as well as swapping out all of it’s marker lights.  Pretty much the rest of the day was wrapped up in half marathon aftermath. I was answering emails from runners asking where their times were, trying to get status updates from Ipico and hashing out a contingency plan.

Mid-day I talked to the Ipico tech guy.  He was making it sound like the odds of them recovering the file were pretty slim.  He also made a Captain Obvious statement that if I would have kept the files from when I scanned all the bibs I would have been ok.  No kidding….  In my email later to my sales rep to open dialogue about what sort of compensation Ipico is going to offer us for this screw up I used this analogy.

To me this this sort of is like taking your car to an oil change place, paying for the service and then having your engine blow up on the way home because they forgot to tighten the oil plug and being told, well, “you could have checked the oil plug to make sure we tightened it”.

That being said, I no doubt am shouldering some of the blame for this happening and will never trust a file from outside source to be legit until I test it.  Later on in the day I got the official word from Ipico that they have exhausted all avenues in trying to recover the chip mapping file, I was on my own.

I had already been working on plan B.  The event photographer, Ken Shelton, had an automatic finish line camera rig on top of a ladder.  It takes a picture every one second.  I contacted him and asked him if he would be willing to upload all roughly 10,000 picture to my home server so I could use them to rebuild results.  He has a vested interest in me getting this done as he uses finish times to correlate with personalized photos that are sent to runners.  He said he was fine with doing that.

My plan was to get all of these files on my network, copy them locally to my SSD drive which will give me lightning quick load times and then flip picture by picture rapid fire to identify each runner and their finish time.  This plan also required me to contact the author of our race timing software.  He made a custom import routine which would allow me to send him a dirt simple file in the format of bibnumber.finishtime, one per line, which would allow me to quickly punch in data exclusively on the number pad.

I started the process myself but things really got cranking when Cindy got home.  I loaded the pictures on her computer and had her be the caller and I was the data entry monkey.  We worked late into the night, getting to almost the 2:19 time in the race which added up to around 1230 people.

This system is not perfect.  If a runner was wearing their bib on their back, on their right side, under clothing or had it obstructed by another runner we could not associate a time with them.  My hope is between the pictures, the manual timing Cindy did on site and the dvr finish line video I posted we can get a reasonably comprehensive set of results.

Now of course this method will not account for chip start times or halfway splits but there isn’t much I can do about that at this point.  I am hoping to finish this tedious process today and be able to post refreshed results this evening.  It’s been quite the journey.  What does not kill you makes you stronger, right?

I shot a brief video of the process last night linked below.

This is how Cindy and I spent the majority of our evening last night.