BT
x Your opinion matters! Please fill in the InfoQ Survey about your reading habits!

Agile 2012: Team Wikispeed
Recorded at:

Interview with Joe Justice and Tom Taber from Team Wikispeed by Shane Hastie on Sep 26, 2012 |
20:40

Bio Joe Justice and Tom Taber are members of Team Wikispeed, a group of volunteers in 18 countries around the world who are working together to "rapidly solve problems for social good" using agile techniques.

The Agile Alliance organizes the Agile series conference, which bring together all the key people in the Agile space to talk about techniques and technologies, attitudes and policies, research and experience, and the management and development sides of agile software development.

   

1. Joe, very briefly would you like to tell us who you are?

Joe: Good morning Shane. Hi, I’m Joe Justice; I’m a business process consultant at SolutionsIQ based in Seattle, Washington and I fly globally; but what brings me here to the Agile Conference is more the work I do after work - nights and weekends; I’m the team lead of Team Wikispeed. Wikispeed is an all-volunteer group that builds a hundred-mile-per-gallon carbon fiber cars; we’re lucky enough to have car #1 behind me, the first car we ever made, which we then competed with in the Progressive Insurance Automotive X Prize, a challenge with a $10M prize purse to see if it was even possible to build 100-mile-per-gallon cars.

Out of 136 cars, we tied for 10th place in the mainstream class coming ahead of some well-known names like MIT, Tata Motors, and Tesla; and I think the majority of how we were able to compete with very large global presences was that we were an Agile team; even though we’re all volunteers working nights and weekends in garages and storage units.

   

2. What drives Team Wikispeed?

Joe: Our mission statement is rapidly solving problems for social good and like many Agile teams when you have incremental visibility into your successes, it continues to motivate the team and there are motivational elements built directly into the process to make people’s success and contributions visible as part of the larger whole; and when it’s a social good initiative or a work worth doing, it’s very easy for the team to remain motivated even as all volunteers. Now we have more than 150 volunteer team members in 18 countries.

   

3. You were recently in New Zealand and I saw you down there so that’s the newest one I believe?

Joe: The newest Team Wikispeed shop is in Christchurch, New Zealand; the city center is currently being rebuilt from the massive earthquakes they had and what that seems to have done is gel the population into understanding how they need to be self-sufficient and supportive of each other and they’re celebrating the successes of doing that with each other and that has made the climate that is a perfect fit for Wikispeed, where you have a mix of altruism and hard work; and an individual in Christchurch volunteered their two-car garage to say “Let’s start building an ultra-efficient car here in New Zealand” and see what Kiwis, the New Zealanders, can do to iterate the efficiency, the ergonomics, the usability, and the safety of that vehicle in one-week sprints like the rest of the team does worldwide.

   

4. So how do you use Agile practices to build a car?

Joe: When we look at the Agile Manifesto, if we substitute the word “software” with “customer visible value”, we suddenly have an elegant set of principles that applies to any industry not just software; and we find that applying exactly those same principles in automotive R&D, automotive rapid prototyping, and automotive manufacture allows us to gain the same benefits of agility that software teams are seeing, higher quality bar, more frequent release, rapid pace of development, and lower cost of development.

   

5. But doesn’t building a car require engineering and all sorts of heavyweight things that software doesn’t?

Joe: What makes it possible is the car is modular; the car splits into eight pieces allowing us to work on 1/8th at a time or more typically all eight in parallel; this is just like software projects; if we didn’t have object-oriented programming, classes, it would be very difficult to achieve the benefits of agility across very large pieces of software; with classes, with interface first design, with contract first design, and object-oriented programming, we’re able to iterate and have code coverage and have test coverage on each part of the project individually and grow it so we’re able to iterate responsibly; we do exactly the same with a car. So it’s predicated on the idea of hardware modularity and that’s loosely coupled modularity where one piece changes and no other part of the car would have to change.

   

6. So the interfaces are well defined but what happens behind those interfaces can evolve and can adapt?

Joe: Right; it’s absolutely black box exactly like you said, Shane.

   

7. And how do you do test-driven development with a motor car?

Joe: So luckily for us the road legal requirements, the NHTSA and IIHS crash test requirements, the EPA and the Unites States Fuel Economy requirements are all written in the form of a test; so we phrase those tests on the wall before we design an iteration of a product and we clearly define how we are going to test those, then we race to build the fastest, leanest, simplest thing that will make that test pass; so we absolutely start with a failing test and lucky for us the road legal requirements, emissions requirements, fuel efficiency requirements, safety requirements are phrased in terms of a test, which absolutely plays to the strengths of Agile, Lean, Scrum, Kanban, Extreme Programming and now Extreme Manufacturing.

   

8. So take me through a test-driven approach to one piece of the motor car.

Joe: Sure; so if we asked a set of team members or the whole team (and people tend to opt in self-organizing team) we need a catalytic converter; that’s the piece of the car, the metal canister with precious metals inside that takes exhaust gas and emits much more clean exhaust gas that you and I are able to breathe safely; if we ask the team to design a better catalytic converter, if they’re a catalytic converter enthusiast, they will build the most gourmet catalytic converter they can think of; they’ll tend to overbuild (in software we would call that gold plating) in a catalytic converter it would quite literally be platinum plating and it would be more than we need; it would probably cost as much as the entire car did the iteration before, just like we see in software. Just like we see in software if we ask someone to solve a problem who does not have a deep concept of that problem, they have to iterate relatively expensively; their first iterations are usually not usable.

If instead we say here’s exhaust and here’s the machine to measure the CO2 content of that exhaust in parts per million, make sure we emit less than 200 grams of CO2 per 100 kilometers driven and we have a test phrase; and we don’t say anything about a catalytic converter; we’re not saying the how, we’re only saying the what; just like Scrum has taught us to do and just like Extreme Manufacturing and Extreme Programming demand that we do for high quality code; when we do that, we then are able to open up the door to any type of innovation and very quick movement because people know right away when they’re done and they don’t have to have any background to begin with; in fact people that don’t are sometimes able to come with the most effective solutions because of the out of the box thinking and the rapidity of knowing when they’re done.

   

9. So in the catalytic converter case, what did you end up with?

Joe: We ended up with two catalytic converters top and bottom integrated directly into the top of the engine, it’s called the head, and where it’s emitting the hottest gas; and so the catalytic converters operate at maximum efficiency when they’re hotter; usually they’re downstream in the exhaust pipe closer to the muffler maybe even behind a resonator; in our case they’re all the way up at the top of the engine where the gasses are the hottest warming the catalytic converter the most quickly activating it and there’s two, a pack of two; and what that’s allowed us to do is not even have exhaust past that point; so we’ve saved about 50 pounds of weight and the engine is run so efficiently we’re not over road legal noise decibels; so we’re able to simplify the system by eliminating the entire exhaust track for they’re back by having an innovative solution where the catalytic converter’s packaged in the head; now ultimately we found a company that was making that solution and now we purchase it from American Honda Motor Company through their Honda racing line.

   

10. So you also buy stuff in where it’s appropriate and don’t just build, don’t build everything from scratch?

Joe: When we make a purchase decision, it’s exactly like a software team would, we wrap those purchase components in an interface so that we then are able to swap it for something modular made by another vendor or made in-house at a low cost of change. So by adopting a small amount of cost upfront to wrap any of these store-bought components, we reduce the cost to make change; the worst thing we would have in team is a slowdown in the pace of development which would happen from a frustration or saying, “But it will take us three weeks to implement that change” – we want to make sure that almost any idea can go from an idea to using and testing that idea in less than a week; we work on one-week sprints on our automotive projects.

   

11. And every sprint produces a production-ready product?

Joe: Yes; it’s all shippable; now that said we’re in prototype; these cars are beta; Ferrari calls this phase owner test drivers; we sell cars now but they don’t have cup holders; their only comfort and convenience acquiescence is an iPhone dock in the dashboard and a stereo. They don’t have heat, they don’t have air conditioning; these cars behind me don’t yet have roofs, they’re all roadsters for quick testing, for getting on road test data of our emissions and fuel efficiency system; so we have beta cars but they are sellable and road legal and some people use them to drive all around.

   

12. And the cars themselves can evolve?

Joe: Absolutely; so people that bring home a car last week, next week, the week after if they wanted to, it could be the last car they ever purchase; the drive train is able to be switched from gasoline to electric making it in a sense “future proof”; or hydrogen or biodiesel or ethanol or methanol because the drive train is a module and it’s able to be switched in about the time you change a tire; the entire car body can switch from the Le Mans race car style car body you see behind me to a pickup truck in about the time it takes to change a tire; so as we have an interior that has cup holders, existing customers are able to swap that on at very low cost in terms of time and dollars.

   

13. One of the missions that you spoke about is doing social good so the 100-mile-per-gallon car is one thing, what else is Team Wikispeed doing?

Joe: The first project I was involved in that was Agile outside of software was polio vaccine distribution to attempt to eradicate polio, that I did with Rotary International; shortly after, Team Wikispeed was growing into its own and we started to adopt projects like that into Team Wikispeed such as low cost simple medical centers for developing communities, such as rotavirus vaccine deployments; the majority of what we do is ultra-efficient transportation; we take on any social good backlog items into our backlog if we have team members that are able to execute on them.

And in fact right now, we’re a little automotive heavy by my eye and I’d like to see more product owners and more Scrum masters and more team members join Team Wikispeed that are passionate about refrigeration, to talk about how to deploy the rotavirus vaccine refrigerated without it expiring to the endpoints where it’s needed the most; or people that are experts in microfinance and help support developing countries with the same perks that first world nations have in terms of investment and banking and savings opportunities; they might be able to do that on a mobile app; many of these countries have mobile data plants but they don’t have stable banking systems; I’d like to see differences made in all of these areas and we’re always encouraging team members to join that have those passions; because right now most of what we do is ultra-efficient transportation which is near and dear to my heart, but I’d like us to be making a difference in the world on all fronts; like Steve Job says “making a ding in the universe”.

   

14. So how do people get involved; if somebody is passionate about solving a social problem; what do they do?

Joe: They send an email to Info@Wikispeed.com and they say “I just heard about you guys; this is what I care about; I’d like to know about joining the team;” then we email them back a team member form and when they read it over, it’s one page; if they decide to sign that’s it, they’re a member of the team, welcome to Team Wikispeed; then we join them into our Google Group where we share most of the data across the 18 countries and most of the conversation happens; then we Skype and we use Scrummy and Dropbox and Microsoft Live Drive, SkyDrive and share information with them wherever they are in the world.

And if they have a garage nearby we then ask them to start looking for the first $488 which is the parts cost to build our entire frame; the vehicle is very inexpensive and very lightweight which is part of why we’re able to have so many people working in so many different countries; many people work from their laptop or from a Starbucks and do phone calls, emails, write music, video editing, project management with us; people that are hands-on typically need a garage or at least a spare bedroom or a closet large enough for one of our modules and they’ll work on it there and then ship it back and forth.

   

15. So you’ve mentioned this concept of the backlog; you have a global backlog of stories?

Joe: We do and in fact it’s visible to anyone in the world at www.Scrummy.com/Wikispeed_all; it’s also editable by anyone in the world; it’s completely open so anyone in the world can add items to our backlog and lucky for me no one yet has decided to delete all the items; it looks like the world actually has a fair amount of social good interest.

   

16. And what are some of the other things that are on the backlog and things that you’ve mentioned the banking, you’ve mentioned a couple of other things but what skills are you looking for and what do people need?

Joe: Right now the thing we need the very most in Team Wikispeed is virtual stress testing of different shapes; so our chassis is, our frame, the insides of the car it’s like an aluminum box and we need to simulate (and we do it now but we need more people to do it) pressure being applied against it or twisting forces so that we can make changes to it and not just rely on the same version; we want to continue iterating our chassis every seven days like we were before and for that we need more virtual stress testing which is often called finite element analysis or FEA; it’s something someone could learn the basics of in weekend if they’re excited or it’s something some people make their whole career out of; that’s a primary issue and one of our blocks right now is we need more of it.

But in terms of skillset, we work with material scientists, project managers, military R&D, tire compound engineers, househusbands, housewives, little kids, girl’s scouts, cub scouts, musicians, artists, knitters all over the world and they’re all able to create meaningful parts of the car and move the project forward and meaningful contributions to our other social good efforts.

   

17. Joe, you’ve given us the link so thank you so much for taking the time to talk to InfoQ today and we look forward to seeing what happens with Team Wikispeed in the future.

Joe: Shane and InfoQ, I absolutely appreciate it; thanks; we’re having a wonderful time at Agile 2012 and we’re excited to during the closing keynote show the current version of our car; I’m glad that people have gotten to see the first car we ever built standing behind me during the week leading up to it; hope to see you at the closing keynote.

   

18. Tom is a Team Wikispeed member based in Las Vegas. So, Tom, tell us about your involvement with Team Wikispeed.

Tom: I’ve known Joe’s sister for many years and I just had coffee with her one day and she said, “My brother is doing this project,” I said that sounds phenomenal, I have to be part of that and two weeks later I was driving my suburban up to Seattle to donate it to the Team Wikispeed cause because they needed a tow vehicle to tow around the trailer for this car; and it’s been about a year since then and I’ve done so many things, learned so much, and had so much fun with the team actually.

   

19. What is your engagement; what are you actually doing with Team Wikispeed and how does it work?

Tom: People ask me that all the time; I do everything from answer our Google voice line; so if you call the Team Wikispeed number it’s either me or Joe that answers the phone, it rings both our cell phones at the same time; I check the email list and just answer questions that people have or ask question; I build 3D models in Google SketchUp of chassis designs or designs for shock absorbers or things like that; I install toilets in the shop bathroom because we had to redesign our whole shop bathroom so our volunteers would have a place to go to the bathroom in our Lynnwood, Seattle shop; so it’s a wide variety; and then sometimes I’m actually turning a wrench on a car; so it’s kind of just a whole variety of things, whatever needs to happen to unblock the current state and get us further ahead in whatever we’re doing.

   

20. So an Agile team, cross-functional, adaptive, just will leap in and do whatever needs to be done?

Tom: That’s kind of how it works around there, yes; just look at what’s happening, what needs to be accomplished for our current goals and our current goals now are big goals are getting three cars built for our first three customers so.

Shane: Because they’ve actually been sold?

Tom: They’ve actually been sold; yes, last week just before this conference, we sold our third car to a local guy in the Seattle area; just heard about us and two days later he was the proud new owner of a Wikispeed car; it’s not built yet but he gave u the check --

Shane: But it’s his.

Tom: Yes.

Shane: And he’s given you the check.

Tom: And he’s also a team member; he and his son are both going to be helping build his car so it’s pretty exciting. so.

   

21. What are some of the really satisfying things that you’ve done over the last year and a bit with the Team Wikispeed?

Tom: It’s been a lot; you know, I think one of the greatest things is actually just getting the car up to the state where it looks really good for this show actually; so we were actually building it to drive it out here and attach some new crush structures on the side, got the interior module set up a little bit better than it was before with some new steering enhancements and seating enhancements; and just when I get that done, doing it with a team spirit and we look back, “Look what we just created with our own hands in a garage just turning tools, we built a car that you can actually drive and it’s like the most efficient car in the world;” it’s just stunning, really it’s just a huge sense of accomplishment whenever I get to work for the team.

   

22. And the car is good; what about other stuff?

Tom: Okay; I think the people that I’ve met in the team as well like some of the people I’ve met and worked with, I just really had a good time learning with them; so there’s that sense of just belonging to something that’s actually changing the world is a huge part of it for me; the chance that I actually get to engineer things without actually being an engineer; now we have a lot of testing so whatever I design or whatever I work on, it has to pass all of our engineering tests and standards; but it allows me to do things that a normal company wouldn’t normally let me do without having been a certified professional engineer but we still have that quality control to make sure that; you know, anybody can submit an idea, anybody can contribute; if you’re an eight-year-old kid and you have a whole new design for the steering suspension which we had a kid that did do this; he said “You know, what if you created this lever and it changed the angle” and we said that’s actually a good idea; we haven’t quite tested it out yet but it’s a really good idea and it was from a nine-year-old kid; what other company in the world can do that; that’s amazing.

Shane: That’s amazing; it’s cool; the motto of doing awesome.

Tom: Yes.

   

23. How does it work?

Tom: It pretty much works using Agile, Lean and Scrum and I’d done some of that in the software world before but never actually in a really functional team that does it really, really well; and having Joe as our team lead he really just kind of nurtures the whole process forward; and having this many people in a worldwide distributed organization is a challenge to kind of get everybody on the right page, head in the right direction and doing useful things but it’s amazing how he just kind of coaches that; and we use Kanban boards, we use a weekly standup call, a global call so everybody calls in team and gets a one-minute standup of what they’ve accomplished that week, what they’re going to accomplish in the next week and any blocks they have; and at the end of the meeting we just start finding ways to unblock the meeting; that’s how I originally donated my car; the block was “we don’t have a vehicle to tow this” and I said “hmm, I’ve got an extra suburban that I don’t know what to do with” and so that became part of the team’s inventory.

General Feedback
Bugs
Advertising
Editorial
InfoQ.com and all content copyright © 2006-2014 C4Media Inc. InfoQ.com hosted at Contegix, the best ISP we've ever worked with.
Privacy policy
BT