Garmin 1000 and Windows Phone Tracking

I started this post a while ago earlier this summer, and it looks like I finished it but never hit post. Oh well! I’m posting it now rather than deleting the draft. After several extra months of riding, my assessment of the Garmin 1000 hasn’t changed, except that I am even more impressed with its stability and ability to record ultra long 24 hour+ rides. None of the earlier Garmins could last that long without crashing!

Ride stats 2

A photo posted by Brian Toone (@kartoone76) on

The positives

  • Large crystal clear display, easy to read in any lighting condition with auto-brightness enabled. In the picture above, compare the Garmin 800 (left) to the Garmin 1000 display (right). Both are displaying 10 pieces of information, but look how much clearer the 1000 is.
  • Di2 integration, no more guessing about whether you are in the 28 or 26 in the back. Plus, being able to see a history of gears used in a KOM effort will be cool to see. Some of the climbs that Mark and I have been trading KOMs on can ultimately come down to using the right gears at the right time. Being able to compare times with the gears used may help optimize KOM efforts.
  • Improved processor speed. The 1000 is much more responsive, able to lookup information on the map much more quickly, and able to calculate turn-by-turn directions for long routes more quickly than the 800. Most importantly, with an improved processor there is presumably less opportunity for deadlock leading to less frozen screens and lost rides. I won’t know this for sure until I get some really long rides done with the 1000. So far my longest ride after the accident has been just over 7 hours. The Garmin 800 was pretty reliable for rides under 8 hours. Once you hit the 8 hour mark, you better not try to do any routing or looking at the map with the 800. You are pretty much guaranteed to have the 800 freeze and possibly lose the entire ride.
  • Improved map detail, plus free maps! Thanks in large part to the OpenStreetMap organization and the USGS, there is some excellent map data that comes standard with the Garmin 1000 instead of having to buy the City Navigator SD card to get good street data. Also, the entire collection of USGS named summits is included in the map data so that you can see a little mountain icon and click on it to see its exact elevation. See screenshot below.
  • The new Personal Records feature is kinda cool, especially if you are coming back from an injury as I am – each time you set a new PR that is especially meaningful.

Garmin 1000, USGS mountain closest to our house.

A photo posted by Brian Toone (@kartoone76) on

The negatives
Let’s start this off with a video showing you the differences in total ascent between the Garmin 800 (left) and the Garmin 1000 (right). Some things to look for in the video – note the iBike gradient stabilizes much faster based on the accelerometer (not barometric pressure). Even so, the Garmin 800 stabilizes much, much quicker than the Garmin 1000. Also, note when total ascent starts to accumulate on the 800 vs the 1000. Note based on the iBike grade how long you are on the hill before the 1000 starts to read any change in total ascent.

ibike data laurel view lapibike data laurel view lap

  • Lack of Windows Phone support. Yes, I know that windows phones are a small (but growing) share of the market, but for people more interested in the camera (Nokia Lumia) than the phone, the current Windows Phone offerings blow away anything from Apple or Samsung. Also, lack of open API for communication between the Garmin Edge 1000 and the Garmin Connect mobile app prevents writing my own home-brewed windows phone app. I contacted support to let them know I was interested in writing a Windows Phone app to support live tracking from the Garmin 1000, but never received any response back.
  • New location of the barometric pressure sensor behind the quarter-twist mount. While this may protect the sensor from rain, it is the nemesis of anyone who lives in an area with lots of steep rolling hills. The new sensor location causes the unit to respond much more slowly to changes in elevation, which means you miss the bottom of every climb. This is also true in large mountain areas, but since you are only doing a few climbs on a ride in the mountains it doesn’t matter much to lose 20-40 feet on each one. But if you are doing hundreds of hills on a ride, then losing 20 feet on each one adds up pretty quickly.
  • The aggressive total ascent filter. It isn’t just the location of the barometric pressure sensor as it is also the “new and improved” algorithm for accounting total ascent. In the Garmin 800, the filter was placed on the raw elevation change. You had to climb more than 7 feet before the elevation changed, at which point every foot was counted towards the total ascent. Now with the Garmin 1000, you see small 1-2 foot changes in elevation, but these are not counted towards total ascent until some magic variable number (I think it is time based) of feet have been climbed. I think you have to have been climbing for 20-30 seconds before it starts to count total ascent, which means that the raw number of feet climbed differs based on how fast you are climbing rewarding slower descending and slower climbing.
  • Total ascent maxing out at 9,999 feet. The Garmin 1000 cannot currently display total ascent greater than 9,999 feet. The display field goes blank and displays four underscore characters instead. (See screenshot below the next bullet)
  • The elevation graph screen only uses the first three sections of the screen. The fourth section is only used if you are following a course with embedded elevation data. This is similar to a bug for the Garmin 800 which only used half the elevation screen. This bug was fixed in an early release of the Edge 800 firmware. The bug currently remains unfixed in version 2.2 of the Garmin 1000 firmware. What’s even worse with the 1000 is that they introduced an auto-scaling feature that scales as if it were using all four sections, but then only uses three sections and chops off exactly 1/4 of your ride. (See screenshot below)

Note the blank “total ascent” field. Also note that the Garmin 800 recorded 12,271 feet of climbing, but the Garmin 1000 only recorded 10,629 feet of climbing. If I had mounted the Garmin 1000 in a normal fashion instead of using the out-front mount with the Garmin angled up at a 60 degree angle to expose the barometric sensor then the elevation difference would have been much, much larger. My lap tests showed a difference of over 40% in total ascent when the Garmin 1000 is mounted in a flat or slightly angled position. Also, note that the elevation graph doesn’t use all four sections no matter how much you zoom in or out. The PR feature is really cool, though, and fortunately it does show you your total ascent – but this is only after you end your ride and only if you actually set a new PR for total ascent. Otherwise, you are not able to see how much you have climbed if you climbed more than 9,999 feet.

Conclusions
Most of my issues with the new Garmin are related to elevation. Other people in the Garmin forums have mentioned random screen locks and dropped bluetooth connections as problems. I haven’t seen either of these problems in the first 500 miles of usage because I am unable to use bluetooth at all. Perhaps the random screen locks are related to bluetooth, which means the lack of Windows Phone support is a blessing in disguise. Overall, I’m happy with the improvements made to the Garmin 1000 and once the elevation readings are sorted out – I’m planning on gluing something to the bottom of the Garmin to funnel air around the mount into the pressure sensor – plus hopefully a new release of the Garmin 1000 firmware will not have such an aggressive total ascent filter.

Windows Phone Tracking

Unable to use the live tracking feature on my Garmin 1000, I decided to write my own for Windows Phone. Without a published API from Garmin, there is no way for me to pull the data coming from the 1000, but I can track position, time, and speed on the phone itself. As I dug further into this, I found an already existing app that looks perfect and works on Windows Phone: locatoweb.com.

LocaToWeb - Laurel View tracking testLocaToWeb – Laurel View tracking test

Everesting Karl Daly

Summit #everesting #fizik #stravaproveit just before the manatees.

A photo posted by Brian Toone (@kartoone76) on

Yesterday, I was able to everest Karl Daly in just under 19 hours doing 57.5 repeats. The exact everesting spot where I hit 29,035 feet of climbing is visible in this pic down near the telephone pole behind me, but I didn’t think it would be right to do so many repeats of Karl Daly without once getting a picture of the Manatee mailbox halfway up the climb. So Chris Shelton took this pic of me next to the manatees dressed up for Thanksgiving. How did I get to this spot? Let’s back up a couple weeks…

Clingman’s Dome Attempt
Each year, my regional chapter of the ACM (ACM Mid-southeast) holds a computer conference for students and professionals in Gatlinburg, Tennessee which sits at the foot of Clingman’s Dome — the highest point in Tennessee and one of the largest climbs on the east coast. I’ve always included some sort of epic ride on Saturday following the conference, but this year I decided I would attempt an everesting of the climb, which would require about 5.25 repeats of the climb. The only problem was an unusual cold spell was dipping the temps down into the low teens. My everesting attempt was on a time schedule (i.e., my wife and I had to be back to Birmingham by a certain time early Saturday evening).

So I started my everesting attempt at 1 in the morning after just 3 hours of sleep. With temps dipping down to 12 degF shortly past Alum Cave, I was struggling with the cold by the top of the 2 hour climb unsure of how I was going to make it back down. It really was far too cold to have even attempted this … my camelbak, food, and my body were all frozen by the bottom of the descent. The logistics of trying to do this in this weather is really crazy. I had to wear my sunglasses in the middle of the night to keep my scarf over my nose, and my hands were so cold I had to do the descent one handed alternating hands with one hand “drafting” behind my back to get it out of the wind. Needless to say, the descent was not only dangerously cold, it was also dangerous! It makes me appreciate the difficulty of climbing everest with all the equipment and trying to keep warm and still access food, drink, etc…

Finishing the festive 500 in the cold and dark.

Let it be said that I’m no stranger to cold (see the picture above from Wisconsin and my blog post about the Rapha Festive 500 competition), but riding slowly through the snow on a mountain bike with hourly stops for coffee, etc… is very different than descending at 30-40 mph for 45 minutes with no place to stop to warm up!

The very next day after returning back to Birmingham, I ran into a good friend of mine Mark Fisher who was out climbing over in Bluff Park. We decided to finish out the ride together during which we tried to figure out the best climb in Birmingham to everest. By the end, it was pretty much decided that Karl Daly would not be the easiest since it was a pure climb with no descents, but it would probably be one of the safest.

Official Karl Daly Everesting
Ben Lowe, the developer of veloviewer.com, partnered with the folks behind the everesting movement and created a map of all the everestings of climbs around the world. He has put together a holding area for riders to submit their everesting attempts, which then get reviewed to see if the attempt followed all the rules. Once the everesting has been vetted, it gets added to the worldwide map along with a page with all the details and comments for the ride. Here’s the link to the updated worldwide map, plus a link to the page created for my everesting of Karl Daly.

Updated everesting hall of fame maphttp://www.everesting.cc/hall-of-fame/

karldalyeveresting on veloviewer.comkarldalyeveresting on veloviewer.com

Statistics
I’ve created a table of all the stats that I find interesting from this endeavor.

Length of climb 2.25 miles (3.6 km)
Round trip (up and down) 4.5 miles (7.2 km)
Vertical ascent (1 climb) 538 feet
Total ascent (58 laps) 29,379 feet (8955 meters)
Total distance (58 climbs, 57 descents) 259.6 miles (417.8 km)
Total time (including stops and complete final climb) 19 hours, 6 minutes
Total time (including stops, to exact everesting spot) 18 hours, 56 minutes
Average speed 13.7 mph (22 kph)
Average heartrate 119 bpm
Average power 162 watts
Average cadence 73 rpm
Average temperature 45 degF (7 degC)
Calories burned 11,839 calories
Total gear shifts 2,718
Total front shifts 130
Total rear shifts 2,588

shifting percentages from my research website di2stats.comshifting percentages from my research website di2stats.com

The Details
The statistics don’t tell the full story, so I’ve saved the longest section for the end — the details. Let’s look at average temperature, for example. Monday was an absolutely beautiful day with temps warming into the mid to upper 60s. It was the perfect day for an everesting. So how did the average temperature end up being 45 degF? Well, I only enjoyed that beautiful weather for about 6 laps (2 hours). Then the temp started plummeting quite rapidly. By 8AM the next morning when I was finishing everything up, the temps had been in the mid 30s for several hours – thus the 45 degF average temperature.

On Monday’s, I bike into work to teach one class at Samford — Software Engineering. After class, I had volunteered to proctor an exam for a colleague, which lasted until 10:15AM. I left work about 10:30 and arrived home at about 11:15, having already ridden 16.8 miles and climbed 2,181 feet. My wife was leaving with the kids for Thanksgiving with her family in Northern Indiana near the border of Michigan. I still had to teach two more classes on Tuesday before the Thanksgiving break, so we decided it would be better for everyone if she got to spend a couple extra days in Indiana while I stayed home, everested a climb, and got some work done during the holiday.

I wanted to see Kristine and the kids off and help them finish getting everything loaded in the car, so I didn’t even leave to ride out to Karl Daly until about 12:15. I took the shortest, flattest route out to Karl Daly since I was carrying 16 pounds of food and equipment and still rode 14.8 miles and climbed 1083 feet to make it all the way to the start. That brought my grand total for the day up to about 30 miles and 3000 feet of climbing even before starting my everesting attempt of Karl Daly.

The weather was absolutely beautiful, though, so I was ready to tackle the everesting fully confident that it would be no problem at all. I ran into Geoff Leonard at the bottom of the climb and stashed my backpack full of stuff behind a tree. He snapped this picture of me as I was taking pictures of my Garmins at the start:

At the start - mile 0, feet climbed 0 - photo by Geoff LeonardAt the start – mile 0, feet climbed 0 – photo by Geoff Leonard

Geoff rode several laps with me, during which time we noticed somebody else doing repeats. It turns out that it was Mike Flowers’ good friend and roommate Matt Finnemore, whose goal was to do enough repeats to climb 10,000 feet in a single ride. We were riding at different paces, but it was great to catch up to him every now and then and see how far he had made it and also relay my own progress. On our second lap together, Louis Pfau arrived with his van that he parked at the bottom. He wanted to try to hit 15,000 feet of repeats while I was doing my everest attempt. Also, at the start of that lap I ran into Terri Jones who dropped off gels for me. Then shortly before or after Geoff left — already my memory is starting to fade — Luke Caldwell made it out there. He rode with me for a few hours climbing close to 5000 feet. During that time, Chris Shelton stopped by and did a couple repeats with the plan to return about 4 in the morning to do more repeats. At the time, I was concerned that I might already be finished by that point. Little did I know how wrong I was!

By the time Luke left to go home with the plan to return around midnight with coffee, it was starting to get dark. I think I did one lap by myself and already was starting to realize that my pace was nowhere near on track to finish by 4 in the morning. I had planned on 16 minute laps, but my time had already slipped to 18+ minutes per lap. Almost immediately after Luke left, Jason Kellen arrived and we rode a few laps together. Sometime during that time, I told him that I was only 75% confident of being able to finish. It was starting to get cold, and I was starting to get tired having barely made it to 3000 meters of climbing with nearly 6000 meters left!

Not too long after Jason left, Mike Flowers arrived and rode with me for a good long stretch. He brought a Starbucks coffee and a big piece of Starbucks banana bread. That jolt of caffeine and calories made a world of difference, and I was already feeling much more confident after only a few laps with Mike. Also, Jason returned with a bag of chips and payday candy bar. I ate the chips on one of the next laps, but saved the payday for later reserving it as an emergency measure should my energy levels really start to go downhill. Mike rode for several hours, and during that time Matt made it up to 10,000 feet. A few laps later, Mike called it a night.

Louis was still going strong towards 15,000 feet and shortly after Mike left, he and I rode together for a lap or two before our paces split us up again. I continued on for a few more laps before Luke made it back with some coffee and ready to ride with me for a couple more hours. Then at about 1AM, Louis reached his limit and decided to head on home leaving his car for me to continue using as a resupply point. By 2AM, Luke was ready to head back home. From 2 until about 4:30, I was by myself battling the increasing cold. My pace had fallen to 20 minutes per lap – over 5 minutes slower than when I first started.

Sometime between 4:30 and 5, Greg Caldwell (Luke’s dad) rolled up and joined me. A lap or two later, Chris Shelton made it back out there. As the sun started to come up, it dropped a few more degrees down into the mid 30s. Greg had to leave for work, but Chris kept rolling with me until the very end when I hit the required 29,035 feet of elevation. I do a lot of climbing on my normal rides, but I favor routes with lots of variety and usually try to find the rolling hills instead of the standalone ones. For this reason, it was easier for me to visualize the elevation gain from above sea level. Several times during the night, I imagined that I was 15,000 and 20,000 feet above sea level looking down far below at where I started. Also helping that visualization, was a set of 10 images/screenshots I made that captured what it looked like at the corresponding elevation on a route from sea level in India all the way up to the top of Mount Everest. I posted these pics to instagram along with a screenshot of my Garmin and a picture of who I was riding with and/or what the Karl Daly climb looked like at that elevation. Here are all the pictures side-by-side (hover over each one for a caption).

What’s next? Well, the goal is Race Across America, and this is one of many adversity training rides where I am stretching myself both mentally and physically to the limit and seeing how well I can respond and continue. More immediately on the calendar — the gravel grovel ultracx series finale race near Bloomington, Indiana on Saturday.

Let me end with a huge thanks to the following people:

  • Louis Pfau – for encouragement and challenging himself to climb 15,000 feet! Also – I owe him a lot for parking his van at the bottom of the climb so I could stash my stuff without worrying about it getting stolen and/or overrun by ants, bugs, etc…
  • Luke Caldwell – Luke came out twice and rode with me for many, many laps including the midnight to 2AM shift when I was in the process of realizing that it was going to take me almost 5 hours longer than I had originally planned.
  • Chris Shelton – Chris also came out twice and rode many laps with me, helping to pass the time and keeping me awake by telling me about his awesome trips to Europe and the climbs there.
  • Mike Flowers – Mike came by with much, much needed coffee early in the evening. I was getting depressed at how slow I was riding, but Mike helped cheer me up — and the coffee provided a much needed caffeine boost. The starbucks banana bread helped me make it through several laps without stopping.
  • Greg Caldwell – Luke’s dad came out to ride several laps with me in the wee dark hours of the morning. He also brought some much needed refreshments.
  • Geoff Leonard – Geoff was there at the very beginning helping to kick things off. Ironically, he lived for many years in the same area as Scott Cole, who is currently just behind me in the climbing competition.
  • Everyone else who came out to ride – Jason Kellen, Rick Swaggler, Terri Jones, Greg Caldwell, and anyone else I have missed. Thank you so much, you all are awesome!!!