Analytics For The Perfect Race
JeanFrancoisPuget 2700028FGP Visits (12446)
Big data analytics proponents keep saying that is can now be used to make better decisions in a whole set of new domains. Here is a great example we have been working on for the past few months. This example also touches topics like Internet Of Things, Mobile, and Cloud.
It all started with a call from Doug Barton, an IBM colleague of mine who is also a triathlete. Doug asked me if optimization technology could help an ultra cyclist named Dave Haase. Dave is set to race the most demanding cycling race ever, the Race Across America (RAAM) for the fourth time. He has been the top US finisher three times.
So far so good, until you know what RAAM actually is. RAAM is a 3000+ mile long race, from Oceanside, CA, to Annapolis, MD. There are no planned stops, racers stop when they want. The first to join the finish line wins. In practice, racers sleep about 2 hours a day, and race the rest of the time.
Add to this that there is over 110k feet positive elevation difference, and you get the insanity of all of this. Yet, people like Dave take the challenge and reach the finish line!
RAAM elevation profile
This is completely incredible.
What's even more incredible to me is that Dave trusted us from the start of the project that we will help him make better decisions during the race. Incredible, but I'm confident we'll really help him.
When I told Doug that optimization can certainly help I didn't realize how much data acquisition and data analysis work this would entail. But here we are, working with a team of IBM colleagues on building optimization models that leverage predicted data using a variety of data sources:
From that data we make predictions about how fast Doug will ride on the road ahead, then we recommend where he should rest. As a matter of fact we are following the general approach I blogged about to deal with big data, see for instance Optimization Is Ready For Big Data: Part 3, Variety
I won't disclose too much before the race on how we will use predictive analytics and optimization to help Dave. Let me just point you to an interview of Dave and I made this week at the IBM
You can find on youtube anot
That's all the details we can provide at this point.
Let me conclude with a short description of the IT system we are building. It touches a lot of hot topics:
This may look like a lot of work, but we are no the ones with the hardest task here. Dave definitely is. We are the ones lucky to work with him.
Update on June 15: The race is about to start. More information can be found on another IBM blog: