Garmin, big data and platypus

shazzz

I haven't seen this posted elsewhere. An interesting tale of Garmin using live analysis of 'big data' to help them win races. Not sure I believe it TBH - sounds like a red herring based on the latest bizbuzz to encourage the other teams to go crazy wasting valuable $s in analysis apps whilst chasing rainbows! Worth a read though.

Oleg Tinkoff also mentioned using big data in his podcast interview this morning but no specifics...

http://www.bicycling.com/tour-de-france/tour-features/oh-geekery?cm_mmc=Bicycling_tdf_NL-_-07242014-_-geekery

above_the_cows

There's Big Data collecting for mining and analysing at a later date and then there's live Big Data analysis. Using it during a race requires the latter and that is very, very, very hard to do.

shazzz

Hard, but not impossible. The company I work for does it, although it is new technology. Well beyond the budget of a bike team I would have thought.
In any case, they seem to be being fairly lenient with their definition of 'big data' at Garmin ...

r0bh

I can't really understand what "big data" was used here. And if their app "accurately reported that Wiggins had been dropped" I might question its accuracy, seeing as Wiggins wasn't even riding!

above_the_cows

I will also say that this is a case of tech being used to present something as new that in fact has always been done before, i.e. basic strategy and tactics, the only new thing is the use of tech in those things, not the actual ability to read a stage/profile/weather report. Plus the example they give is not proof of anything. They won the stage, but they can't say it's because of this platypus.

I am trying to calm my inner ANT analyst.

above_the_cows

r0bh wrote:

I can't really understand what "big data" was used here. And if their app "accurately reported that Wiggins had been dropped" I might question its accuracy, seeing as Wiggins wasn't even riding!

I'm thinking they might mean Titchy Richie, but then my big data telly box accurately told me that Tithcy Richie had been dropped. As I'm sure did race radio, the Tour de France website, twitter and all other manner of social media. The only data they would not have been able to mine was this forum as it crashed that day.

iainf72

I guess, with enough data and the right algorithms you could work out pretty quickly if breaks will work, when to attack etc etc

Anyway, I can't think of it as big data unless they have a significantly huge amount of data. And I just don't think there is enough of that from cycling yet.

jawooga

• complex model sitting on a database +
• big data +
• streaming 2-way cross the world in real time between mobile appliances.

As the cliche goes, take 2 from 3. Probably very impressive but I suspect it's not big data in the current meaning (which to be fair is pretty open to interpretation, and is as much of a buzz word as anything).

Does anyone know what database (or type of DB) they're using? Don't think it said there.

mroli

Above The Cows wrote:

r0bh wrote:

I can't really understand what "big data" was used here. And if their app "accurately reported that Wiggins had been dropped" I might question its accuracy, seeing as Wiggins wasn't even riding!

I'm thinking they might mean Titchy Richie, but then my big data telly box accurately told me that Tithcy Richie had been dropped. As I'm sure did race radio, the Tour de France website, twitter and all other manner of social media. The only data they would not have been able to mine was this forum as it crashed that day.

I think the article is talking about Wiggins being dropped in the 2013 Giro?