One Lap Around Lake Travis 2013: April 27

November 27th, 2012

Mormon Mills climb

The Spring ride season is right around the corner (that’s what I keep telling myself), and plans are underway for the 2013 One Lap Around Lake Travis. We’re planning on 6:45am, April 27th. The route is not yet final, it might be a little longer than last years 120 miles, but we’re planning on finishing at Jester King Brewery again.

A little Pave

April 10th, 2012

Did you watch the Paris Roubaix on Sunday & think “I’d like to ride 160 miles on the worst roads I can find?” Well, I’m not going to tell you that ride exists in Central Texas, but I can point you to a 120 mile ride with a 3 mile dirt section. Instead of pedestrians crowding the roads, you’ll find cattle and goats. Instead of hoisting a pave above your head, you can lift a pint of Jester King’s Wychmaker. Sold? Join us on the One Lap Around Lake Travis on May 12th.

New route planner

February 12th, 2012

Big news! For last couple of years we have been working on some underlying technology and lately we (“we” meaning Bryce) put in a big effort to get it working. The preview router planner is now live! Why does the world need another site that generates routing directions? Aren’t Google and Bing Maps and all of the others sufficient? In a word, No. Not for cyclists. At least, not for us.

True, Google Maps does support cycling directions, but have you actually tried using them? In practice they aren’t very good. The recommended route into town from my area goes up Far West Blvd, a 20% grade climb, and through a high-traffic intersection. Far West is a great climb if that’s what you are looking for, and the intersection isn’t too bad for experienced commuters, but there are considerably easier, and flatter, routes if you know the area.

My commute goes where?!

The issue isn’t with Google or Microsoft, it’s with the lack of information about the roads. Cities and counties publish lists of recommended bike routes and expressways are flagged but many other factors go into determining whether a road is good for cyclists or not. Does it go through a high-traffic commercial area? Are there stop signs every block? How are the shoulders? What is the grade of the hills? How the heck are we going to come up with this data if Google and Microsoft can’t???

We aren’t – all of us are. By aggregating the uploaded GPS data we can begin to see where cyclists really ride. Combing publicly available data sources such as Open Street Maps and NASA’s SRTM elevation data with the GPS data and our customized data mining and routing algorithms lets the router tart to make routing suggestions better suited to the needs of cyclists. For example, the total distance is often less important than the quality of the roads. There are lots cool ideas we have in the pipeline…

Big dreams, but we aren’t there yet. The route planner is just a preview of what’s to come. It still makes silly suggestions about riding down interstate highways and there are lots of data problems out in the hill country that result in some humorous routes. However, getting all of the infrastructure and data models in place to be able to do this basic routing quickly is a big piece of the puzzle. We can now start iterating on the more interesting pieces. Stay tuned!

Bikes, BBQ & Beer 2012

January 23rd, 2012

We’ll be doing the 3rd annual version of the Bikes, BBQ & Beer ride on March 17-18. If you’ve never done this ride before, it is 120-130 miles around the Texas hill country. The ride starts and ends at Coopers BBQ in Llano, with an overnight stay in Fredricksburg (and a visit to the brewery). The ride is free, but you still need to register.

Register here

Time to plan for the spring

October 7th, 2011

Time to start thinking about planning some organized rides for the spring. I’m thinking about mixing it up this year.

The annual Llano to New Braunfels BBQ ride is still on, though it might change to simplify the logistics. The ride is always a lot of fun, but finishing the ride 150 miles from the start makes managing transportation a pain. On top of that, the start and finish are both an hour drive from Austin. Day 1 was amazing last year, so we’ll probably stick to the roads around there, throwing in some more hills, cows and water crossings.

We’ve also got another ride in the planning stages that should be a good one. You’re gonna need your climbing legs and chamois cream. Anyone know someone at a central TX brewery willing to sponsor the ride, preferably west of Austin?
hills

cows

New Bikko Profile Features

September 16th, 2011

The profile pages have gotten a good workover. Gone is the ugly long list of rides, and in its place are new stats and an activity calendar. The stats plots are still under construction, and will evolve as we discover what users want to see. If there is something you’d like to see, let us know.

Goodbye wind, hello hills!

August 20th, 2011

My family and I completed our move back “home” to Austin from Wichita, KS this summer.  I say “home” because neither of us are from Austin, but given the number of family and friends here and the Austin culture it is hard to tell sometimes.  As cycling cites go they are about as different as they get, though that isn’t a knock on either.

Wichita is certainly not a leader when it comes to bike friendly cities.  There are a couple of bikeways (meaning literally two that I know of) for commuting and just in the last year got the first two bike lanes painted onto First and Second streets going into/out of downtown.  Rural roads rarely if ever have a shoulder, let alone one that is wide and clean enough to make a difference.  That said, I was very pleasantly surprised at how generally considerate Wichita drivers were when passing. There was the typical, occasional unnecessary horn from an “aggrieved” motorist but no more than Austin or other cities and very few close passes.

As far as cycling goes, it is hard to described Kansas as anything but flat and windy.  All of my rides tended to either look like sticks or rectangles because even out in the boonies everything is amazingly straight (http://bikko.net/route/561 or http://bikko.net/route/434).  And that means even a short ride may entail 10 or 20 miles into a 15+ mph headwind.  In theory going into a headwind is little different than climbing a long hill, but I haven’t lived anywhere I can consistently climb for 10 miles without a break!

Hills on the other hand I find more gratifying.  Not only is there a clear end to the pain but there is usually a nice view to add to the sense of accomplishment. West Austin and for many miles further to the west has an abundance of rolling hills. We found a house within 15 miles of downtown, with many bike lanes to choose from, so I will have one less excuse to prevent me from riding to work a couple times a week.  Of course, we also managed to pick a location that ends my commute with a grueling 380 foot climb. (http://bikko.net/route/671) Oh well, it is great to be “home” and if anything at least there is a view at the top of each climb.

Webservice API details

August 11th, 2011

While doing the visualization talk I got a question about how to get data out of bikko. I haven’t got around to publishing an API doc yet, but here is the details on how to get raw data out.

Find the URL of the route you’re interested in. For example, the URL for my May crash is http://bikko.net/route/572. To get the raw data, do a HTTP GET with HTTP_ACCEPT=application/json in your header. The content type application/trift is also supported if you’re concerned about bandwidth and/or serialization speed. The response will look something like this:
{
"center": [30.3785, -97.7935],
"zoom": 13,
"distances": [0.0, 1.0, 2.0, 3.0],
"speed": [0.0, 2.0, 4.0, 6.0],
"power":[0.0, 200.0, 220.0, 240.0],
"altitudes":[735.0, 735.1, 735.2, 735.3],
"latitudes":[30.3785, 30.3786, 30.3787, 30.3788],
"longitudes":[-97.7935, -97.7936, -97.7937, -97.7938]
}

If there is any other data you’d like to get via the web service API let me know.

Visualizing cycling data with Python

August 11th, 2011

Last night I gave a brief talk on how I used Python to visualize the data I collected during my May crash. The goal of the visualization was to determine the speed at which I hit the curb, which required required using OpenStreetMap, Google Maps, Chaco, and a bit of Python. I also demoed some of the other tech I’ve worked on, such as hill detection, though those demos didn’t make it into the slides. Here are the slides and the source code

The day my helmet saved me

May 6th, 2011

I had just climbed a few hundred feet, turned around, and was enjoying a nice descent. I was cruising at 43 mph, with the steepest part yet to come: a 22% grade hill sweeping to the right. Then it happened. I heard a ping, and immediately the front wheel became unstable. A spoke had just snapped on my front wheel. I was on the brakes immediately, but because of the grade, I wasn’t slowing. The bike started shaking and I could no longer steer. I wasn’t going to make the turn, and was heading towards a curb, guardrail and canyon, in that order. The bike was bucking like a bull, the handlebars jumping and shaking. I pulled the handlebars to the right, then bars bounced back to the left, this time more than I could control, dumping me on my left side hard. After that, I’m not entirely sure what happened. I know there was sliding, maybe a tumble, then slamming into the curb. My head snapped downward, hitting the curb.

Many X-rays and a couple of cat scans later, the verdict is no serious injuries. My left side is pretty banged up, a puncture to my left knee, and a very small fracture to the base of my skull. My helmet though, is trashed. The front left side has a serious crack, above the left ear, another crack, and on the back, three tiers were cracked. The helmet saved my head. Thanks Giro.

If anyone is interested in the route details, my gps was a little off, I didn’t actually fly into the canyon.