Am I the only one concerned that Google’s latest attempts to improve search involve creating artificial intelligence? Amir Efrati has previously written about this on the Wall Street Journal’s blog, and we get periodic reminders that AI is still a full-steam ahead, top secret project at Google, and that it’ll be the Next Big Thing in search algorithms. The justification, as Efrati’s article describes it, is a better ability to comprehend the vagaries of human language.
“Jaguar” is the example cited in the article–only a thinking computer can figure out if you’re referring to an animal or a car. Scary as sentient computers might sound, what’s particularly cool about this is that language no longer has to matter. Finally, Google’s AI will let us sever ties to the cumbersome roots of words and base their meaning on whatever’s trending these days.
Sadly, last night left me little time for bike design, as it involved a meeting about said bike and the complex machinations that are calculating box dimensions so that Cyclocross.com will be able to ship complete bikes to happy mudlovers without incurring the oversize shipping wrath of UPS. What I need, really, is a computer capable of human thought. And a self-riding bicycle to get me to work.
According to BoyGeniusReport.com, Woz is predicting cloud computing tragedy as the world slowly realizes they no longer own anything–that they’ve basically ceded ownership of all of their digital stuff.
While I’m sure Woz knows more than most of us, I have to wonder what it was we used to own anyway. Apparently I’ve already drunk the Cloud Kool-Aid, but I can’t think of anything I now store online that would’ve been better confined to a single hard drive. Maybe it’s just that the pace of things has accelerated to the point where nothing we “have” is of any value any more.
“Ownership” has been on my mind since the Danzig patent and the various chaos surrounding patents in the bike industry these days. Intellectual real estate certainly seems to be tightening up in some industries. I’ve been storing backups of all my Solidworks files online, and I still think of those as mine. But even though I’m pretty fond of this suspension system right now–sort of all-consuming–I can still see it as an evolution. What it is right now, isn’t what it’ll be in the future.
I think that’s the weirdest thing about designing something. Whether it’s a little computer code or this bike frame, you have to be completely invested in it. At the same time, you have to realize it’s only a moment in time. Put a little more simply, for me the hardest thing about designing something is stopping.
Case in point: in the process of reshaping the crankcase that houses the lower link, I started to notice a few more things I could do with the lower link. I’ve lowered the pivot locations again, and shortened the link, forcing a reset of the system that I believe–at the other end of a bunch of additional hours–will be better. Here’s where it sat as of last night.
I’m pretty committed to the idea of the even shorter lower link, and I’m liking what it lets me do with everything else. That’s what I’m working on right now.
Uneventful day here at Canootervalve right now while I’m at work on another revision. At least the temperature’s finally breaking some. This past weekend we headed to the coast to escape some freakish triple-digit temperatures in Portland and Vancouver. Everybody else had the same idea, but we finally found a little piece of beach and some 66-degree temperatures. One nice thing about the Pacific Northwest is that you can almost always find a completely different temperature within a 90-minute drive of wherever you are. I’m new to beaches, but I can’t imagine taking a taxi there is particularly common.
The next round of Danzig revisons is all about the “crankcase,” that machined lower pivot housing. Unfortunately, I need some quiet time under the hood on this next phase of that, and won’t have anything to report until the next round of drawings are done. More photos as soon as they’re ready to share.
A little intense, this living bikes 365 right now, but fun. Wake up, ride to work, obsess about cyclocross and e-commerce, ride home, obsess about NoTubes, then put a half hour into frame design. Lots of time to think on the rides back and forth.
Great meeting yesterday about the future of Danzig, too. There are a lot of ideas in the air, and that’s the best part.
The biggest design challenge ahead for prototyping Project Danzig is the piece I’ve dubbed the “crankcase.” It isn’t the busiest pivot location I’ve seen, but I still want this piece to be as clean and simple as possible.
To that end I keep refining this section, and probably always will. This most recent iteration is an attempt to make things even simpler and lighter. When it comes to Solidworks, I’m ten levels below “amateur,” but I’m constantly blown away by how quickly Solidworks lets you create. Here’s the quick 2D redraw of the crankcase.
And just exiting the drawing recreates the extruded 3D piece.
It’s pretty incredible, how quickly revisions can be made to pieces. The question is how streamlined can this piece possibly become. That’s what I’m working on this week.
I’ve lost track of the hours I’ve put into Project Danzig, and probably best I did. But just when you think you’ve designed something pretty cool, a company in Japan comes along and releases giant robot that shoots people when you smile.
Lest that shot past you, let me just clarify that last sentence. One feature of the giant new armed robot being sold by Japanese killer robot makers, Suidobashi Heavy Industry–I mean aside from the fact that you can control it with an iPad–is that it shoots people in the face when you smile. Smiling operates the dual Gatling guns that fire BBs, so the funnier and funnier you find it that you’re using a giant robot to shoot your friend in the face, the more Kuratas shoots him in the face.
What type of insurance does a company like this buy?
Did I mention you can operate it from your smartphone? Of course you can.
Nothing I write about can possibly compare to the promotional video, though. Until you’ve watched an adorable Japanese girl’s demure giggles cause a robot to shoot someone in the face, you’ve never met the 21st Century.
My suspension design, in contrast, might one day make people smile, but that smiling won’t be able to put anyone’s eye out. At least not yet.
Clearly I have a lot of design work to do.
To that end, today we wrap up axle path–or the “versus propulsion” side of it. Here’s the drawing from yesterday.
The red dots mark three key stages of the axle path. If you look at the green vertical line along which the red dots are arranged, you’ll notice the lowest dot–which represents the bike sitting there with no weight on it–begins on the right hand side of the vertical line. As the bike moves up to its sag point, the axle moves to the left of the vertical line. The key is that the axle path continues to move rearward for 10mm of vertical travel beyond the sag point.
The idea is that chain tension continues to counteract the suspension’s movement until you’re 10-15mm beyond the sag point. The sag point is where the suspension stays when you’re riding around. Chain force hinders suspension movement until you’re 10-15mm beyond the sag point. Once you compress the suspension beyond that point chain tension no longer tries to control the suspension.
So that’s the idea. Almost all suspension bikes work this way, so I’ve not created anything unique there. What I did try to do, though, is develop a platform that will allow a huge range of options with all of these points. The system makes it easy to fine tune every aspect of the axle path.
So that’s what we’re going to do.
Responses to my “How much more would you pay for U.S.-made carbon fiber?” question keep coming in, but at this point the clear winner is $250-$500 more. The majority of respondents would pay up to $500 more for a carbon fiber frame if it were made in the U.S. Impossible to say how accurate that admittedly highly unscientific analysis really is, those are the results. I’ll keep an eye on incoming responses, though, because these things can change.
And maybe next week we ask how much more you’d pay for a giant killer robot that was made in the U.S. Makes me smile just thinking about it.
BikesComments Off on Short-listing Adam Craig for a Pulitzer then Back to Axle Paths
Aug022012
That photo’s from Adam Craig’s blog. He just posted about his inaugural Megavalanche race–that insane DH race with an eight-thousand plus foot descent, 350 rider mass start on snow, and genuine chance of dying. A lot of my work involves reading, writing, and patching up code for posts. I end up reading a lot, and Adam’s posts always stand out. You have to like phrases like, “I overtook dozens of people, mostly in a safe manner,” and “a 4m wide ‘road’ made up of fist to lawnmower-sized rocks.” Most professional riders lead incredibly interesting lives, but the ones capable of actually describing those lives are few and far between. Adam’s one of those great riders who can write about it.
Anyway, Danzig: so I’m thinking the best possible scenario for version 1 is this: 29er, 125-130mm of travel, 17.35-inch chainstays.
Version 2? Well, that’s something else entirely.
I suspect the last time I tried to articulate what was going on with the axle path made little to no sense, so I’ve been meaning to double back to that for some more detail. For starters, here’s the general axle path of Danzig:
Here’s that same deal, only zoomed in and with highlights on the sag line (where the suspension should be when the rider’s weight is on the bike) and dots to emphasize the horizontal orientation of the axle relative to that sag line and a hypothetical straight vertical line. I came damn close to using “vis a vis” instead of “relative to,” but I fought it off there.
What the hell’s all that, then? At this point, most of you probably already know, but let’s wrap it up with a final detailed explanation of chain and axle path tomorrow.
Most people in the bike business have come to accept the, um, eccentricities, of the supply chain. Rick Vosper’s done a brilliant job of breaking it all down at BicycleRetailer.com. It’s a very good read–particularly if read along with any of the recent articles about U.S. production and carbon fiber.
Most of these articles, like like one on Bikerumor.com, mention Intense. Brands like Intense and Turner have become the quintessential examples of U.S. manufacturers at a crossroads. Stateside production of a competitively-priced carbon fiber bicycle frame isn’t an option, so they’re weighing the loss of control that comes from overseas production with the fact that nobody will buy a frame that’s much more expensive than comparable models from other brands. And increasingly few customers will pay a premium for an aluminum frame.
The 30,000 view of these two issues is pretty interesting, as is the recent “high tech but made in the U.S.” marketing push from brands like Enve. The trend at this point definitely suggests we’re reaching a kind of tipping point with high-end bicycle frame manufacturing. The desire to create high-end product at something like a competitive price–and manufacture it in the U.S.–is out there in a way that it didn’t seem to be even a year ago.
This will probably never happen–not because we can’t do it, but because so much of bicycle industry culture is deeply invested in Asia. It’s the current path of least resistance. There’s a hell of a lot of resistance, but still less than trying to make carbon fiber bike frames in the States.
Obviously, this is much on my mind right now, as has been for some time.
So I figured I’d ask you, my distinguished panel of authorities. Assuming all features of the bike are comparable, how much more would you be willing to pay for a carbon fiber frame to be made in the U.S.?
You thoughts are much appreciated, and I’ll report back the findings in a bit. I’ll also be back to Project Danzig next time.
How great was Marianne Vos’s Olympic Gold? First Wiggins hauls some major sideburn action into Paris and then Vos takes a brutal-looking rain-soaked women’s race in–if I’m not mistaken–a skinsuit. It was tough to explain to the kids why I was waving a cowbell at the TV.
It’s been one hell of a busy few weeks. The 2013 NoTubes catalog looks to finally be ready to head to print. I’d written parts of that all across the country, including a Holiday Inn laundry room. Cyclocross.com, on the other hand, is just kicking into high gear in anticipation of a launch date that already can’t happen fast enough. It’s all been sufficiently panic-inducing to pull a guy off his pet personal project, which by necessity has to get whatever time I have left at the end of the day start of the next day.
Still, I was able to do a little reworking during my last round of PF30 conversion revisions. I wondered how short I could possible get the chainstays.
Now, there are a million and a half ways to cheat on chainstay length. My favorite is probably cranking up the bottom bracket height, but in Danzig’s case, that wasn’t an option. So the question becomes: how short can I get those chainstays for a bike with 29-inch wheels, a 12.77-inch bottom bracket height (can always go higher, but that’s as low as I’d go) and (let’s just figure) 130mm of travel. With some adjustments I went ahead and made, looks like Danzig might be able to drop chainstay length down below 17.4-inches.
That’s pretty short. It’s not a done deal, yet, but it’s close enough for me to say that’s now my target.
BikesComments Off on Thanks to Sugar Wheel Works and Digging into Axle Path
Jul302012
Thanks to Jude at Sugar Wheel Works, my old Independent Fabrications single-speed is slowly turning into a bike again. When NoTubes sent me some rims to help bring an old Industry Nine wheelset back from the dead, I was really looking forward to asking Jude to do the build. There’s just a great vibe to Sugar Wheel Works, which shares an enormous building with a bunch of other really interesting businesses. In addition to building wheels, Sugar Wheel Works also offers wheelbuilding classes. It’s a simple kind of business model–offer a service, plus classes on how to do it yourself–that seems to be growing. Only today I found out Carl Schlemowitz from Vicious Cycles is starting something similar with frame building and painting.
So the Indy Fab is on the way back to life. Now it’s down to sorting through boxes and trying to find the rest of the parts I need to finish piecing it together.
This weekend I also finished a pivot redesign on Project Danzig that relocated all the pivot points to allow for even more freedom of motion around the bottom bracket shell. The goal was to ensure ample room for PressFit 30 and similar bottom bracket diameters. In the process, I had to rebuild my main frame around the fresh pivot locations and recreate my axle path from the ground up (literally). Here’s an approximation of what it looks like right now.
We’ve looked at instant centers, axle paths and chains a bit, but one of the specific elements I’ve prioritized throughout all the iterations of this design has do with the relationship of the axle path to the chain. The drawing above illustrates my goal for the axle path.
How you arrange your pivots dictates not only how far rearward your axle can travel relative to its starting point, but at what point in the compression the axle is furthest away from the bike’s bottom bracket. As the axle travels upward, it has the opportunity to move rearward slightly. What I want with this design is for the point at which the axle is furthest away from the bottom bracket to be just past the sag point on the bike.
Why is that? Because chains aren’t elastic. Tension through the line of the chain will keep trying to pull the rear axle toward the bottom bracket shell. By designing a suspension system that needs to move rearward in order to keep compressing right at the point at which pedaling tries to pull the rear axle forward to counteract that rearward movement, the obvious goal is to cancel out pedaling forces. We’re talking halves of millimeters in some cases, but the objective is to turn the chain force into a positive when it comes to eliminating pedal bob.
Factor in the much larger wheels of a 29er or 650b bike, and it can become more challenging to align those two. It was a need to create a longer travel 29er that still had short chainstays that led me to design a frame in the first place. Going forward, I’ll try to explain a little more about the challenges of a bike with larger wheels, how axle path gets involved in all this, and how that led me to design this system.
BikesComments Off on Monkey-approved Kids’ Bikes and Danzig PF30
Jul272012
I have to come up with some bikes for my kids. Actually the girl child has a pretty badass little 24-inch Santa Cruz I concocted based on an idea I got from framebuilder and ultimate bike-for-the-kids-creator Jeff Jones. The boys need at least one new bike, though, and I wish I could find something that looks pretty much exactly like that monkey’s bike.
I’m serious. Why doesn’t anybody make like a kids bike with larger wheels, but small crank arms–like 145mm? That’s sort of what I did with the Santa Cruz, and it makes perfect sense. Bikes like that are ultra-stable thanks to the whole sitting “in” the bike with some bigger wheels up around you (you’re basically creating a scaled down 29er in that regard), but everything else is kid-size. So there’s that on the agenda, if I ever get my garage finished.
I spent most of yesterday writing product descriptions, put in what felt like my weakest ride all year to get home, wrestled Facebook in a professional capacity (just like wrestling it for fun, except for the company issued luchador mask), then wrote some javascript to help make a site compatible with iPads. That last one is positively insane, as I haven’t written any genuine javascript functions in over a year. It was a little like meeting an old friend who no longer speaks the same language, and trying to build an airplane together out of shit you found behind a Radio Shack.
Oh, what Javascript and I had been through together! And to think now we barely know one another. I’m so happy with it that way.
Anyway, I seem to’ve found a little extra in the tank to crank out this Danzig redesign. I’m happy to report that after obsessing about pivot locations on my ride into work this morning, and on my sorry excuse for a ride home, it occurred to me that I can generate plenty of space for a PF30, or a PF50, or whatever the hell massive oversized beast becomes the next standard.
The nice thing about the relocation is that it also makes the frame easier to build. The key was a proportional move up and back for the lower pivot. Here’s Danzig PF30.
Remember yesterday when we looked at instant centers and how their vertical position affected axle path? Maybe it didn’t sound like that yesterday, but the bit about the higher the instant center the move rearward the axle path would be–that stuff? It turns out that axle path is my guiding principle when I’m moving my pivot locations. Right or wrong, there’s a set of criteria I’m following that dictates my axle path, and that range is what dictates my pivot locations. Using more goofy graphics and animations, I plan to dig into that next week.
