Joining Bobby and Jens this week is a genius who changed cycling as we know it with his radical approach to science. His legacy includes snub nose time trial helmets, specialist time trial suits and away from cycling - Eliud Kipchoge's sub two hour marathon.

Remember you can ask your questions for our next mailbag episode - wherever you see @oddTandem... or you can email them to us oddtandem@shockedgiraffe.com

Join our Patreon now to get an ad free version of the podcast and your name in the titles

https://patreon.com/OddTandem

Find us on social media:

https://x.com/oddtandem

https://facebook.com/profile.php?id=100093419004559

https://tiktok.com/@oddtandem

https://www.instagram.com/oddtandem/

Hello everyone, and welcome back

to another episode of The Odd Tandem.

2

Yes, you know that we've changed our name

because you're listening.

3

But, make sure to spread the word

to your friends that we're back Jensie.

4

We are back and we try to be back

bigger and better.

5

So we have more episodes, per month

with some extra features.

6

We still have super interesting guests.

7

We would still talk about everything

related to cycling, but

8

we just try to be better

and more fascinating than we were before.

9

So yes, please share the love

10

and share knowledge with your friends

about The Odd Tandem Bobby.

11

And yes,

12

and that we are.

13

We are an odd tandem

aren't we Jensie you know.

14

No kidding.

15

Been friends a long time

but yep, you know,

16

I think you're on the front of the tandem.

17

I'm on the back because, you know,

18

you were always that guy in the front

going in breakaways with that game face.

19

And I was always kind of crystal

cranking in the back.

20

So if we ever get some pod art made,

you know, it's got to be you on the front

21

with your game face and me on the back.

22

Just, I don't know, reading the book or,

eating some carbohydrate hits

23

in the background.

24

But, anyway,

so today we have a fantastic guest.

25

you may not know his name, but anyone

that's been involved in cycling and

26

especially sports science, data science

for the last decade plus knows his name.

27

So give a big odd tandem.

28

Welcome to Robbie Ketchell.

29

All right everyone special treat today

Robbie Kettle.

30

Welcome to the odd tandem.

31

thank you so much for having me.

32

And I love the name, by the way.

33

Yeah.

34

Thank you very much. I mean,

35

we went through quite a few names,

there's no doubt about it.

36

but,

37

who would you have on the back of your odd

tandem?

38

Do you have anyone in your life

that, like, ends

39

and I that are just totally opposite,

but it just kind of clicks for

40

for you in that person?

41

Oh, well,

42

I obviously have my son because he's, he's

the one that actually pushes me to move.

43

So, you know, otherwise

I wouldn't be on the bike.

44

I probably have him with me.

45

How how old is he?

46

How hard can you push you

or you have to carry him around?

47

Or is he, like,

is he your little turbo engine?

48

He's six, but he's a superhero.

49

You know?

50

So he gets that he he does.

51

You can fire that thing up. You know.

52

Yeah.

53

you move the camera a little bit,

but let's see the shirt.

54

The, Oh, yeah. That's. Yeah.

55

Baby.

56

It, we just had a walk for a while, so.

57

All right, you know Wyatt.

58

Wyatt has down syndrome.

59

And, my wife put on started

the first ever Seacoast

60

Best Buddies friendship walk here

in, Exeter, New Hampshire.

61

And so this was the second year running.

62

And so Wyatt has his own team.

63

It's called Wyatt's Warriors.

64

well, we just have fundraising

for Best Buddies, and,

65

we just got,

we just got done with that on Saturday.

66

Cool, cool.

67

Well, you know, to our listeners

that may not know your name,

68

I mean, you are an absolute legend

in the sport of cycling.

69

You kind of started the whole sports

science phase.

70

so tell us in our listeners

71

a little bit

about your educational background

72

and how you got involved with the sport

of cycling in the first place.

73

Robbie.

74

oh. I actually start a little bit on my,

passion for the sport.

75

So I grew up a bike racer.

76

I wanted to be like you guys.

77

I, started bike racing

when I was 12 years old

78

because I blew out my knee

79

and I couldn't do any of the fun sports

like soccer and football,

80

because back then, they didn't repair

the ACL on a 12 year old.

81

and my parents bought me a bike.

82

and that year, when I was 12,

83

my dad drove me from upstate New York

to Wichita Falls, Texas.

84

I think it was like a three day

drive to go do the National

85

Junior National Championships.

86

And I sucked so bad

that I got dropped on the first lap.

87

So my dad drove me three days

to come to this thing, and I just got

88

absolutely shot out the back

of a bunch of 11 and 12 year olds.

89

but from that day, I was hooked.

90

And, I raced my bike until I was about,

91

25 years old,

trying to trying to make it a career.

92

I think I once made $500 in a year

racing my bike.

93

I mean, I, I did my best,

but I it just wasn't it wasn't my thing.

94

it wasn't meant to be, I guess.

95

And so that was always the connection

with cycling.

96

I was, computer science and, engineer.

97

and had, then working, in a lab

98

in, Colorado State

in, Fort Collins, and,

99

there was, so I was in grad school

100

there and there was, a wind tunnel,

and there was a training company,

101

and two of them came together,

102

and the wind tunnel had nothing to do

with cycling or sports.

103

And the training company

basically wanted to

104

I think it was called Colorado

Premier Training at the time.

105

They wanted to do

wind tunnel testing there because

106

wind tunnel testing was just starting

to become like the cool thing.

107

San Diego.

108

There was a wind tunnel out in San Diego.

109

and, there were a bunch in Europe

110

and, you know, they were creating these

cycling tests balance.

111

And, and so I got to be

112

part of, a team that built the software

113

and the testing balance for that,

114

wind tunnel in Colorado

115

and Fort Collins and the same to see,

you know,

116

was like the most retro

wind tunnel of all time.

117

It was made out of plywood.

118

It it looked like a cardboard box.

119

because it wasn't

meant to be like a formula one type of

120

sexy aero type of thing.

121

It was used for gas dispersion modeling,

and so nobody really went in there.

122

They would run the wind tunnel

123

like all night

on these little tiny buildings to look at.

124

Like, you know, how from health science

buildings, the fumes,

125

whether they went into a sidewalk

and like things like that.

126

And we built

a, a balance that went in this thing

127

and the balance set off the floor

by like four feet.

128

So you had to, like,

you know, rock climb yourself up

129

to this thing,

and to get on the, on the testing balance.

130

And it was looking back at it

like it was the least safe thing

131

that you could put,

like any person on at all time.

132

And the wind tunnel didn't

turn off either.

133

So like in between testing,

you literally the wind, the wind's

134

going 30 miles an hour, 50 K an hour,

and you would have to take a bike

135

with a disc wheel out the door

136

perpendicular to the wind,

you know, and hold on to this thing to go

137

and, like, mount it up there.

138

And long story short,

it was Slipstream Sports.

139

That was, I believe it was Garmin

Slipstream at the time.

140

The name of the team.

141

And, they were based out of Boulder,

which was an hour south of us,

142

and they booked 100 hours in the wind

tunnel that year.

143

And that's how I got connected.

144

back in dead days,

145

my very first,

wind kind of testing was in East Germany.

146

I was just 18 years old, like,

maybe a year before the wall came down

147

and they used,

148

this stuff.

149

They made candles of paraffin. Paraffin?

150

Is that the word they would, like?

151

Heat it up

so it would become smoke and dead.

152

Yeah.

It's like a boil based. It's an oil base.

153

Yeah, right. Except you

you did the same thing.

154

So you could, like, visually visualize

the air flow, it seems.

155

Yes, yes, we did it.

156

Except for the things that we had were,

they used that all the time for the gas

157

dispersion model, because that's

what they would do to simulate the flow.

158

They would take videos of like flow

coming out of exhaust from buildings

159

and these buildings.

160

By the way, there were a couple inches

tall type of thing.

161

And they would, and

so you would look at the flow over there.

162

And so we had these

like retrofitted things that were meant

163

for the gas dispersion modeling,

164

and we would put it over to the cyclists,

and you could see the flow over.

165

And, you know, most of the time

we would like, run out of things

166

to test with people. So we would be like,

look at the smoke.

167

We could put it around and show

you like how it is and try to educate.

168

you know,

how how the how aerodynamics where it.

169

Yeah. Okay.

170

So that was one of the things that I had

on my list, but you're kind of responsible

171

or maybe fully responsible

for quite a few other innovations.

172

And I had to lean on Christian Vande

Velde our Buddy Christian,

173

and, he kind of went down a list.

174

what other products or technology

175

did you help

kind of bring into the world of cycling?

176

that is pretty, pretty standard.

177

Now, I have two that I know of the,

178

the helmet, the Giro helmet that you made

179

evidently pulled an all nighter

at training camp and, like, filed down

180

the longtail to be like, oh,

the only Christian knows, story that's.

181

Look, that is now,

you know, pretty much a norm.

182

And, the old famous, what was it called?

183

The bat suit or the squirrel suit?

184

The time trial suit?

185

Yeah.

186

Tell us how you came up with these ideas.

187

I mean,

it was over a decade and a half ago,

188

but you were kind of,

like, on the right track, where now it's.

189

It's just the norm.

190

Yeah, it's funny,

I forgot about that story with Christian.

191

So just to give a little bit

of background to this, so when I started

192

working in the wind tunnel,

this is before I was hired, by the team,

193

we were working with the team,

but they were working.

194

They came to our wind tunnel and,

195

I basically had free rein to go in

and run

196

the wind tunnel

in the middle of the night,

197

because remember

what kept the lights on it?

198

That wind tunnel was doing all this

gas dispersion modeling.

199

And so doing the cycling testing

was kind of an inconvenience to them,

200

because they would have to change out

the balances and everything.

201

And so it was like

it I'd have a key to the door.

202

It was like 5:00 that the normal staff

would leave for the day.

203

And I would go in to this warehouse

and I would have to turn the big breaker,

204

switch down and turn on the turn on the

fan, and like, the wind would come up and,

205

and I would be able

to study things in there.

206

And I had like foam

207

that I would sand down

with like sandpaper.

208

I had like these foam cutters

that's like this hot wire.

209

And I would cut like different shapes.

210

And I was making things like,

that went, at the bottom of, the shoes,

211

that went around the,

the cleats and like, the, the pedals.

212

I was testing, like, different helmet,

shapes and stuff like that.

213

And so I had

I was fortunate to have, like, a resource

214

where I had endless amount of time

215

in the middle of the night, like,

I could go all the way until 9 a.m.

216

if I wanted to, with just running the wind

tunnel and testing things.

217

And by the way, there was a lot of foam

that just kind of flew back into the sand

218

because a lot of stuff

got lost during that time.

219

but it was at a training camp

with Christian,

220

I think, Christian had broken

his collarbone, in the Giro that year.

221

We were trying to get him ready for it.

222

So I fast forward.

223

I was hired by the team away

from the wind tunnel.

224

And this is like the first training camp

that I ever went to.

225

So the team, like all the resources

were at the Giro still,

226

and they were like, okay,

we need to put, Christian at altitude or,

227

you know, somewhere where he can recover,

but also kind of get back into the swing

228

of things and like,

hey, Robbie do you want to go,

229

you know, drive the car behind them

and, like, support him and do things.

230

And so I had so much time to watch

Christian ride his bike

231

and, you know, see him in his TT position

and just be like, hey, we should

232

we should be doing some narrow stuff

while we're here.

233

And so I took his TT helmet.

234

I went to the warehouse, the,

like, hardware

235

store and, bought a hacksaw.

236

got some of, I think

237

it's called Fapose foam over in Spain.

238

And,

cut off the tail of his TT helmet and,

239

you know, sanded down, the, the,

it put on a glued on this big foam

240

block and descended it down,

just like Arts and Craft project.

241

And he would go ride the next day in that,

and he'd be like, how do I look?

242

You know, and I just get some feedback.

243

And that's how that kind of

244

t helmet kind of came, came to place,

245

it was that year

246

that we were Pearl Izumi was a sponsor,

247

and there was an intern at Pearl Izumi.

248

They used to come up

and spend time in the wind

249

tunnel with me in the middle of the night,

and we were trying to design a,

250

new speed suit for the tour,

that we could do.

251

And we went pretty overboard

with the wings underneath the,

252

the armpits that year, to so much

that it kind of restricted

253

the guys from putting their arms out

in certain positions.

254

I think there's a picture of,

Ryder Hesjedal where

255

his not really in his position

because he was, over 40 K.

256

It was like pulling it

257

back, quite a bit, but, yeah, that's,

that's how that's how those came about.

258

So, there

259

for, for my understanding

and our listeners and viewers,

260

at the start, you said you were into

a data analyzes and computers.

261

How did you make the transition to become

an aerodynamics expert, self-teaching.

262

Or you did more studies, to cover

debt base

263

or how how did you change from

being in an office at the computer

264

to the wind tunnel and be, like, doing

practical work out there in the field?

265

Yeah. Great question.

266

I mean, obviously

aerodynamics was cycling

267

was something that I knew a lot about

because I was a cyclist.

268

Right.

269

And we're all aware

of the effect of aerodynamics.

270

I had studied fluid dynamics, in, in,

271

in college and, and, and gone

through the engineering courses for that.

272

those subjects, however,

273

had never done wind tunnel testing,

274

until we had built that, that when,

275

when tunnel testing balance in Colorado

and it was

276

I would say through hundreds of hours

of, of

277

just kind of being able

to study aerodynamics

278

as it relates to a cyclist, that it became

kind of a specialty, for me.

279

and, you know, it became a lab.

280

It became like,

you know, a lot of times the scientists

281

like you go to a new lab and you're

studying in new concepts and stuff,

282

and the same scientific principles

apply in the sense that you need to set up

283

your your, your different tests

that you're going to do.

284

How are you going to test?

285

and then you, you know,

back into the science around that.

286

and keep in mind that aerodynamics

is a pretty heavy computational, field.

287

there is, computational fluid

dynamics has been around,

288

and that's what we used in, you know,

the marathon project more recently?

289

completely.

290

We we barely did any wind tunnel testing,

and that's all computer simulation.

291

And, and that's kind of

what I'm working on now,

292

are some alternatives to aero testing,

because

293

keep in mind, like wind tunnel testing,

as you guys know, is not, is not it's

294

a big inconvenience to bring an athlete

to a wind tunnel, especially when you're

295

trying to do altitude camps and you're

trying to do all kinds of other things.

296

And that's what led to the bat box.

297

Bobby, I don't know

if that was on the list by Christian.

298

but I built, hand soldered together,

299

dynamic pressure sensors

that were like pedal tubes, essentially

300

that, where, on a circuit board

where the microprocessor

301

there that measured the wind speed

and wind direction on a bike.

302

So these things didn't exist.

303

You see, these sensors are out there now.

304

And this was back in 20 2012.

305

I believe that it created the bat box

and the bat box, set on a bike.

306

And the reason that I created that

created it for guys like Christian and,

307

and David Millar

and all the guys on the slipstream team

308

at the time, because it was so hard

to get them to a wind tunnel.

309

And the reason I bring

that up, Jens, is because,

310

you know, the wind tunnel

and the aerodynamics

311

is one thing around it, but it's there's

so many challenges to being able

312

to kind of study that on, on a cyclist

313

and have it applied to specifically,

you know, our, our sport.

314

In the period that you were

315

starting to think like this

and you were coming up

316

with these amazing ideas, a lot of

317

it had to have been

318

trial and error.

319

Oh, yeah.

320

so, you know,

we've talked about some things that are,

321

you know, entrenched in, in cycling now,

but what are some of those things,

322

maybe those projects that didn't quite

get off the ground

323

that you were working on way back then?

324

There had to have been some crazy idea

that you had that.

325

Yeah,

I had gotten into the professional teams

326

through being an expert in aerodynamics.

327

However, because it was

328

so logistically challenging

to do aerodynamic testing,

329

there was so much limited time

that we actually did the aero testing.

330

So I had to apply other things in life

331

things around biomechanics,

things around physiology, things around

332

nutrition and stuff, and kind of,

you know, be a sports scientist.

333

and data science was, data scientist

was not even really a term at that time.

334

Right.

335

And, I was really keen on,

336

being able to come up with some,

analytics platforms,

337

and nobody was really doing

true analytics.

338

And I wasn't even like something

that you heard.

339

And I created a software application

called platypus

340

and platypus was to, as far as I know,

341

the first AI system that any pro team

used in the in the peloton.

342

And this,

343

this, application,

344

was essentially

345

I had to solve for the fact

that when you're like team Sky

346

or your CSC, like when you guys

had the yellow jersey in the tour,

347

like you always had the front most cover

position, right?

348

And like you, you knew more

what was going on in the race,

349

especially back

then when radio communication.

350

I don't even know

if it's any better today,

351

but like, it's always a problem, right?

352

It's always a problem.

353

The caravan

and but if you're like 20th car,

354

like 15th car,

you have no idea what's going on.

355

Like you really have no idea

what's going on because you,

356

you you,

you're not in the race a lot of times.

357

Like, I had the luxury of being able

to drive the car every once in a while

358

in smaller races.

359

Sometimes you just kind of go up there

like, oh,

360

I thought they thought we had a rider back

here, you know,

361

just because you're bored.

362

and what,

363

what I did was, I created this

application that was, gathering source

364

that Twitter, was starting to be,

365

become really popular, at this time.

366

And I created this application

that was scraping Twitter data.

367

And we basically would,

368

gather all the information

that was mentioned about our riders.

369

And then I would take it and I had this

truth like validity algorithm around it,

370

and it would basically try

to indicate an and, an appropriate time,

371

like, hey,

did we have a rider that flatted,

372

so that they could hear it before or hey,

do we have riders that are

373

hanging at the back of the peloton

374

and don't look like

at the bottom of a climb,

375

like they're starting to sag

or something like that?

376

And, it gave the, the directors.

377

Charly Wegelius

378

was the one that was using it at the time,

gave them more information,

379

and it didn't completely remove the,

the radio, but it was supplemental.

380

But the challenge was

381

the connectivity of the service,

382

like, for Wi-Fi and cellular networks

were such a challenge

383

that there was limited time

that I could send information to them.

384

And so I would have, like, preprogramed

like information

385

that would, be,

so I would only send like a signal of,

386

like the number one,

and that would tell them rider at the back

387

or that would tell them, like,

388

you know, we're all at the front or

something like that, because the smaller

389

the data packet, the easier

it was to send it to them at the time.

390

And so we had to we had to.

391

So now you have things like Starlink and

like all these different like satellite,

392

internet, services

393

that would probably make an application

like that super useful.

394

but that's something that like we tried

395

and was hard to not really take it.

396

And, and the riders got pretty frustrated,

I think, because I was

397

I was driving ahead of the race

to go to the hotel

398

where I could watch the TV too,

by the way, and be able

399

to also add information into

this and send it to them.

400

but instead of being on

the side of the road, like handing bottles

401

or like doing something

that was like more urgent,

402

maybe like that they, so

and so it didn't take off.

403

And I think, it, in today's age,

they have more of an opportunity.

404

But the,

cellular networks were the limiter there.

405

another example of something

that was a success

406

but almost wasn't a success.

407

if you look at those,

those stubby helmets that came out

408

and, it was 2011 that we introduced them,

409

we had put a big focus on,

trying to win the team time trial to tour.

410

We were like, we had never won

a stage in the tour with slipstream.

411

And we were like,

all right, team time trial.

412

Like nobody focuses on the team time

trial.

413

Let's go win that.

414

And we didn't take Johan Vansummeren.

415

And he had won Paris-Roubaix

to the tour that year.

416

And that was kind of

that was a pretty big deal

417

that we didn't make that decision,

you know, and we had taken,

418

it was we had taken some young riders

419

that, were kind of, people that even

the media was like, really?

420

This is going to be the team

we had to push off.

421

we had Christian, we had David Millar,

we had Dave Zabriskie.

422

So we had like a

pretty like, top hitter team, you know,

423

and we, so we bring these helmets there.

424

We had never used them yet, because

we're trying to keep them kind of secret.

425

And we,

426

I laugh about this

because now the ability to create, like,

427

prototypes is so much greater than it was

428

But, like, we bring this to, the race,

429

and they had never really worn them

before, which is always a mistake.

430

Right.

431

But we also we're kind of up against

a timeline to, to bring them there.

432

And we put them on the riders to go

433

do like the, the pre ride the course.

434

Christian comes back to me

and he throws a helmet at me

435

and he goes, that helmet sucks.

I can't wear it like it.

436

The shell came disconnected from like the

the protective layer inside.

437

It was like going over his eyes

and he couldn't see.

438

And he goes,

just give me the other helmet.

439

for for the team time trial.

440

And, you know, keep in mind

this is like an hour,

441

hour and a half until, start time.

442

And, and I looked at him on the side.

443

And, he didn't know,

but we had sent the old helmets

444

back to the service course that morning,

so the old helmets weren't there.

445

Right?

446

So I go I go into the truck

and I don't tell them this

447

because I'm like,

I gotta go find a solution, right?

448

I can't go with them to him

with the problem.

449

And he's at this point, he's on them.

450

He's on the trainer,

like warming up with the rest of the team.

451

And he's in.

452

He's seen me walk by like a million times

and he's like, you got the other helmet?

453

And I'm like, And so I, I like

454

go into the truck, the mechanics truck,

they get some superglue.

455

Right.

And I'm trying to glue this thing on.

456

And then it's me and one of the,

this one years.

457

They're trying to like you know, hand

drive the thing.

458

Couldn't even find a fan.

459

And we're like, okay, cool.

460

We gotta go show them that we fixed it.

461

So we go to give it to them,

462

super glue and dry it.

463

So he puts it on, his eyes burn, and he's

just like crying inside this helmet.

464

He takes it off and now he's getting

a little irritated for for good.

465

For good reason. Right.

466

And and so he's like,

give me the old helmet.

467

And I'm like, So I go into the truck

and I'm like, okay, last chance.

468

He was a small helmet and we had a medium,

469

and I bring him the medium and I'm like,

hey, try this on.

470

This is probably a good idea.

471

And he puts it on.

472

He's like, oh yeah, I guess I'm a medium.

473

And that was like five minutes

before the start.

474

And they like roll off to the

to go to the start.

475

And I was driving third car that day.

476

So the team time trial

you can have three cars right.

477

At least used to be I don't know anymore.

478

But and we had had the plan

that Julian Dean

479

was going to pull off

after the first kilometer.

480

that was just like the game plan.

481

Like he was just going to go hard

right out of the gate,

482

and he was going to pull off

and just drop, drop off.

483

It was a short time trial

thinking about like 20 K or something.

484

So like there was no issue

with the time limits.

485

So he pulls off in the third car.

486

That means that like I just follow Julian.

487

And so I have no idea

what's going on. The race.

488

I think I messed everything up

with Christian.

489

Like we put everything

into the whole season for this

490

and I have no radio contact.

491

I'm so far away from everybody

that I'm like,

492

I have no connectivity to anything,

so I don't know how we're doing.

493

I get, we get back,

I park the car, Christian pulls up to me

494

before I even get out of the car,

and he goes, never do that to me again.

495

It's like, you don't really know

what happened there, but I got it right.

496

It, so I get out

and see that we're winning, and,

497

we had to wait for, Sky to, to finish.

498

I think they were last off

or close to last off.

499

And we had beaten by one second

to win the team time trial.

500

And that was the team's first,

stage win in the tour.

501

Wow. That is pretty dramatic, right?

502

You couldn't

write a script more exciting than that.

503

That's really awesome.

504

So now that you talked

about all these ideas and how cycling

505

has changed in terms of aerodynamics,

for example, do you

506

where or how do you see the next

big jump or development?

507

I mean, for the last let's

say 5 to 8 years, maybe it was nutrition.

508

So what would be

the next field of development?

509

Tires maturing after tire again?

510

dynamics, the weight of the bikes,

the shape of the frames or

511

where can we go further and faster now?

512

Yeah.

513

You know, I truly believe that.

514

It's like the circle of life that as you

push one field, like nutrition,

515

it actually opens up areas with training

516

or opens up areas with bike position

and things that you didn't realize before.

517

And then it starts

circling back to the thing.

518

Oh, hey, we weren't actually doing that

the most optimal way.

519

I think training science with,

AI and analytics has has really big,

520

opportunity.

521

I don't think anybody's

truly doing it yet.

522

the platform, it's so hard to build a

platform to do it right, to provide value.

523

I think that, analytics,

advanced analytics with,

524

AI in all kinds of different fields

aerodynamics, training, nutrition,

525

you know, recovery,

526

like all that's

just going to start to really boom.

527

the equipment is so interesting

because it has to follow the,

528

the ceiling that the UCI gives it

from the regulations.

529

And that's changed so much. Right?

530

I mean, you look at how advanced

531

and now that it's like

we went back in time for a number of years

532

with the way that the frames

look with the geometry and,

533

you look at like Bjarne Riis's bike,

right?

534

When, when he won the tour with that

535

Pinarello like,

that was really, really advanced stuff.

536

And, and now you see,

with the handlebar positions, like,

537

what's kind of coming from that

now that they're able

538

to have them at a specific angle.

539

And they weren't allowed to do that for,

for a number of years.

540

and so position is always going to evolve

because of that stuff.

541

the things that people don't

think about, like,

542

all the bearing frictions

and the different technology

543

that come out with that stuff

and how much of an impact

544

that really has over time

for, for, special pro cyclists.

545

so it doesn't really answer your question,

546

but I just think that it's going

to be discontinued to, to cycle through,

547

like all these different fields

and there'll be continuous, innovation.

548

So basically, you

549

see, we developed this just pulls up

550

and we focus on that, that pulls up

a third part, we focus back on that.

551

So it builds on top of each other.

552

Yeah exactly.

553

Yeah it does.

554

Yep I see I see.

555

And a lot of times

we think that like we're doing something

556

completely right

557

and we can't

558

can't get more advanced with training or

we can't get more advanced with nutrition.

559

But then we're like,

oh, wait, we've been doing it all wrong.

560

And something that we've done

in a different fuel kind of unlocks at.

561

Not very often

do we have someone like you,

562

who's always the smartest guy

in the room on our podcast.

563

So I want to pick your brain a little bit

about thermal regulation, about ice baths

564

recovery and heat acclimation

because again,

565

you, you know, Christian told me that

566

you guys were thinking about that

sort of stuff way back in the day.

567

So kind of

from when you started to where it is now.

568

What is your opinion of

of those two things in particular?

569

Yeah, yeah, I had a pretty it was a pretty

hot topic for a number of years.

570

Allen Lim, who has,

571

basically

had my job before me at slipstream,

572

he, he started, pushing

573

pretty heavily on that,

and with things like the ice fast,

574

and when I got to the team,

we were still considering,

575

different ways for thermal regulation,

especially for recovery.

576

However, just like any field,

like you find something that advances it,

577

like the ice fast, and you realize

how inconvenient the ice fast is and how,

578

like, hard it is to like, truly prepare,

especially when you have team time trials

579

and you got to have like eight ice baths

for people, right, for the entire team.

580

And, we, we looked into,

581

just, ice gloves or using slushies,

582

which is probably the most effective way

for cooling core body temperature.

583

because you actually lose most of the,

584

the heat and the energy

through, through your urine.

585

when it leaves your body and,

however, there were challenges with that

586

because when we put a slushy

machine in the, in the bus,

587

the bus heated

588

up because it was so hot

from running that thing inside the bus.

589

but we literally had, like,

a slushy machine that you would see,

590

unlike,

with an ice cream truck or something.

591

that was, that was on, on the bus.

592

And it was so hard to resource that,

that we only had like, one good one.

593

And so there was never one

on the second bus.

594

And stuff like that, all things that you

just think would be so easy to achieve.

595

And maybe in today's age, like there's,

you know, more ways to resource things.

596

but then we,

we started, freezing tights, when,

597

when Ryder won, Ryder Hesjedal won the,

598

Giro 2012, we were doing ice tights.

599

And so when he came on the bus

after a stage,

600

we would have, like,

leg warmers and arm warmers

601

that had been soaked in water

and then put it in a freezer.

602

And he would immediately

put that, on his legs and,

603

and it stripped down

and put all the stuff,

604

and it would lower his core body temperature and it would accelerate the recovery.

605

you know,

that's the recovery piece of it.

606

and it really, you know, he was somebody

that adopted that technology, too.

607

Like, it really takes a special individual

608

to have that commitment that after,

you know, you're absolutely cooked,

609

you're getting on the bus,

you're hungry, media wants to talk to you

610

and you're like, no, I gotta go

put my ice tights on for for five minutes.

611

And the, and he really committed

to that, that year.

612

Now, the,

613

in terms of,

like, thermoregulation, like,

614

we're getting a little bit outside,

a cycling, but,

615

you don't realize when you're drafting

like, it, it really affects

616

your thermoregulation because you don't

have the wind over you to cool you.

617

Right.

618

And there's a really there's a, there's

619

there's a lot to be said

about the position that you ride

620

and the clothing that you wear and,

how your body cools.

621

You look at, Pogacar.

622

during the Giro, he saw climbing.

623

He had tights on.

624

He had leg warmers on. Right.

625

And he, didn't have anything in his arms.

626

And he had something over his, his, neck.

627

And that's the way

his body cools effectively.

628

Right?

629

He was able to regulate is temperature

really effectively

630

at like, high intensities

with those colder temperatures.

631

By dressing like that,

I don't think we dress appropriately.

632

I think he's really on to something there.

633

And, a lot of times you see people

wearing arm warmers, not wearing

634

leg warmers the other way around,

but the legs are

635

the ones that are propelling

you up the mountain.

636

Right.

637

And, and so you have to

you have to manage, like, not just,

638

when you say core body temperature,

but you need, you need to address, like,

639

you know, the,

the different, muscles in the,

640

you know, the legs versus the arms

and the entire body in the face, like

641

how and how that effectively cools you.

642

and the the

643

thing that I wanted to bring up was,

when we did the 159 project,

644

in Vienna with Eliud Kipchoge

645

for the sub two hour marathon.

646

well, we didn't realize

647

we had created this,

648

like, incredible pacing formation

where it really it literally created

649

aerodynamic lift for him

when he was running,

650

when he was running behind these, runners

in front of him.

651

And so it was

it was sucking him along, essentially.

652

And what we didn't realize

so is that affected his ability to cool.

653

And we had to because we didn't

get the start time that we wanted

654

because they wanted to have when we're

going to have crowds and everything.

655

And we wanted to start at 5 a.m.

when it was going to be cooler,

656

we had to tell him

657

to use it as a tool

so that if he was overheating,

658

to get out of the pocket

of where he was running

659

so that he could, you know, maintain

his core body temperature.

660

And so drafting is a big, like a big thing

about,

661

regulating your, your temperature.

662

And of times you can't choose like,

if you're, if you're just holding on

663

for your life on the peloton and you're

drafting like you're going to draft,

664

but you may not understand that, like, I'm

actually overheating and I need more

665

wind on me or,

you know, something to be able

666

to manage that effort.

667

yeah, I think I lost a little bit

of what the question was,

668

but there's, there's lots of ways

to thermoregulation for sure.

669

I actually believe Formula

One cars have the same problem.

670

They keep getting told,

go out of the slipstream, get cold air

671

and go back in a slipstream

672

because the engine just exactly

as you said, the engines are overheating.

673

so now I have a question for you as well.

674

Also coming from formula one

675

and Formula

one decode basically was all the data

676

collected from their own team

and what they know about other teams.

677

They can say okay.

678

Yes, I believe his tires

going to last three and a half more laps,

679

and then he's going to do

refilling for 7.8 seconds.

680

We is a push now or stop now.

681

So now in cycling, do you believe

682

sophisticated or teams was a bigger budget

and having resources.

683

Do you have a computer program

like for example, Pogacar, his team,

684

do they have a computer program

that says, okay,

685

Jonas

Vingegaard can push 431W for 18.7 minutes

686

if we can hold the same wattage

for 19 minutes, we beat him.

687

You think we are down to that? That you.

688

No. Precisely.

689

Looking at the other rider,

his body weight, his position,

690

the energy output,

how he can beat him, in theory at least.

691

Do you think we have programs like that

692

So right after the 1:59

I went back to working for teams.

693

what was Ineos Grenadiers then?

694

we created

695

software that was called

Competitor Awareness.

696

And we essentially we're trying

to get to that.

697

The challenge

is that you don't have the actual

698

formula

one that data is is more commonly shared.

699

so like the metrics on the engines

and like things like that

700

cycling,

it's really hard to get power data.

701

It's kind of it's becoming

like a little bit more commonplace.

702

So you have some data,

but you don't have like years

703

of training data

and some of it's on Strava.

704

We were scraping Strava data.

705

We were doing all kinds of stuff

on like the riders.

706

But then I wrote, code

that essentially was,

707

determining the watts

708

per kilogram for a rider going uphill,

for all the, competitors.

709

Because if you

710

going uphill is easier to try to calculate

711

because drafting has less than a fact,

because you're going at slower speeds.

712

And if you know where they're starting

from the video like of the race,

713

and you know what the pitch of the road

is because your riders actually

714

went on the same road, right?

715

So you actually have that GPS data,

or you had the

716

the data before you even went to the race,

and you had a rough estimate

717

of what the rider's weight is.

718

Then you can determine what the watts

per kilogram is,

719

because you can calculate it from that,

and then you can determine way over

720

what distances they can do that

on what pitches, they can do that, etc..

721

So I think that there are ways to come up

with kind of niche metrics,

722

but holistically

you're still kind of like,

723

not able to get all the data,

from specific riders.

724

Now, keep in mind, like riders change

teams though, exactly what have you.

725

My next question. Yes. Yeah. And ethically

726

that data belongs

727

to that with GDPR laws and everything.

728

That data belongs to the rider.

729

And however

you've developed models off of that rider.

730

And so like you have like

some pretty abstract things that,

731

you know

now like you've learned about that rider.

732

And so, you know,

you look at, Richard Carepaz

733

and you know, that was at Sky and then,

734

you know, went to EF

and they know everything about him,

735

you know, they know so much about him,

and not just

736

from what his power numbers are and stuff,

737

but they know how he performs in the cold,

even without having numbers on it.

738

They just know from experience, like,

739

oh, this is the type of climb

that he won't perform

740

well in, or this is this type of finish

that he's really going to perform well.

741

And because it's punchy

and he has a good sprint,

742

you know, things like that,

and you learn, a lot.

743

I have to say that,

like, cycling is unique in the sense that

744

everybody like,

745

thinks that it's so scientific.

746

There's so much art to it still.

747

Yeah, there's so much I mean, the,

748

the directors are so knowledgeable,

749

you know, they're bike

racers have done it for 20 years.

750

You know, they've known, like most guys

in the peloton, they still know of raced

751

with like, you can't you

you can supplement that.

752

But that knowledge is so, so important,

you know.

753

And that's why, you know,

a computer doesn't direct a race.

754

the director still does.

755

And it's why I, I will not take a software

engineers job.

756

Right. Because there's still the ability

757

you still have to be able

to direct the tools.

758

And so I think,

759

we get lost in the fact that we think

that, like, technology is free.

760

These guys have these, you know,

helmets on that, looks like spaceships.

761

with these bike that came out,

those Giro helmets this year and like,

762

they, these. Why?

763

What now? We're going to have windshields

in front of the bike and everything.

764

Everybody makes all these comments,

but they don't realize that

765

when you have a specialized tool like

that, you also need to use it effectively.

766

And those riders, like the engineers,

spent years developing it.

767

And then the riders,

768

are they have to learn how to use it

effectively.

769

I watch them riding them like they're

they're not using it.

770

Right. Like they're going to and they're

going to get big gains out of it.

771

But their heads up like half the time

because they're trying to use a helmet

772

that's specific for going fast

and then aero position at high speeds.

773

And they're using it on a Prologue course

that has 50 turns in it.

774

And they got to look up, you know.

775

And so it's like it's it's it's

hard to apply, the analytics that you're

776

talking about with like the competitor

awareness, like all that stuff.

777

There's so much art in it still,

778

man, I'm glad that you said that because,

779

you know, I look back at,

780

the data science that we had back when

we were racing and I wish we had more,

781

but then I'm looking at all

the data points

782

that are available to guys

and gals out there now.

783

And I kind of feel sorry for them

that there's so many data streams

784

that they have to think about,

because what you just said is, is spot on.

785

the the riders have to learn

how to use it and use it correctly.

786

And that takes a specific type of rider,

787

like you mentioned before, somebody

that's very open to that.

788

And if you're just piling this

on these younger

789

and younger pros because let's face it,

they're coming into the peloton

790

at 18 now and expected to be winning

pretty much right away.

791

You know, this gives such

like a neural load to these poor kids

792

that there's so many things

that they have to think about that

793

it just makes me a little bit nervous

that that

794

that artistic side

that you also mentioned, you know, that

795

that tactical side that, that, you know,

796

put it all on the line

side is going to be affected by,

797

by some of this data

because, you know, a rider,

798

you know,

certain riders are racers, right?

799

Like they go when they feel like going.

800

And we've seen that a lot with.

801

Yeah.

802

Philippe Gilbert

was one of the best examples of that.

803

today

Pogacar is is obviously a prime example.

804

but like if you're having that artistic

805

tactical flow being overridden

by a person in the car saying, oh,

806

this sensor says this

or this calculation says this, or it

807

is it taking the fun out of,

808

of our sport a little bit.

809

Yeah.

810

I mean, to be honest,

I think it makes it more exciting.

811

I think that we

812

continue to perfect our craft

813

by having these tools that are available.

814

And some of the tools also keep in mind,

make the sport safer, right?

815

Or have the potential

to make the sport safer.

816

and listen, I'm biased

817

because I'm on the data guy

and I want to continue

818

to innovate in the sport

and and see more of this stuff.

819

be developed.

820

But to me, it makes it more exciting

because there's there's more things

821

to talk about.

822

There's more things that the general

cyclist can, can connect with.

823

Because if they're if they're seeing data

talked about about,

824

say, aerodynamics or power to drag

or things like that,

825

there's an opportunity

for them to go and experience it,

826

you know, and there's,

because if it was just, about,

827

riding the tour de France, like,

they're not they're never going

828

to experience that unless, you know,

maybe they could go ride some of the,

829

some of the finishing climbs

and, and circuits and stuff like that.

830

But it's, it kind of to me, it adds

another level of excitement.

831

But keep in mind that there will always

should be

832

some art around it

and how you use those tools.

833

just like any, any,

834

any profession really.

835

so, no, we talked about too much science

836

in cycling, but you were also,

a part in a very scientific project

837

in, what track and field

are running the 1.59 project?

838

so you were part of a team

that, broke the two hour,

839

timeline, right?

840

And, but just for our viewers

to understand better,

841

in order to run a marathon in two hours

or less, that's 21.

842

Isn't our average?

843

Honestly, Bobby and me,

we struggled to keep that on a bike.

844

Average. Right, Bobby? 21.

845

You, us average for two hours.

846

I got gotta work hard to make it

so let's fill us in a little bit.

847

okay.

848

Did the three we're able to run a 2:05 a

2:04 beforehand, but what did you change?

849

How did you make it run

just a little bit faster.

850

Yeah.

851

So the previous

852

and Eliud Kipchoge

who is the runner who went under two.

853

he held that record.

854

And this the first attempt was in 2017.

855

Nike had decided to go for,

it was it's always been

856

a benchmark that people have been trying

to achieve, like the four minute mile.

857

And, when,

858

so Nike

859

decided to put together a project

860

I was really fortunate to get invited

to be a part of that project.

861

There was five of us of the core

kind of scientists and engineers that,

862

that got to, to work on that.

863

we put together, like,

where are we going to do in the world?

864

what should the course look like?

865

What should the weather conditions be?

866

How what should the patient strategy be?

867

how can we run off of,

you know, with other pacers

868

and runners to help

with aerodynamics and drag the clothing?

869

Should they wear? She. Should they wear.

870

What's the nutrition

871

and all that same type of things

that you try to optimize for a bike race

872

only we got to choose

the location, the time. And

873

the location

874

at the time, that we wanted to do it.

875

So in 2017, we missed, it by 26 seconds.

876

So he finished.

877

Eliud finished

in, two hours and 25 seconds.

878

I never thought I was going to be part

of a project again.

879

like that

880

2019 rolls around.

881

I had left the sport of cycling

and running completely.

882

I was working on autonomous vehicles.

883

I never thought I was going to be back,

working on anything.

884

And, and endurance sports

885

and, I'm in the I'm in the nick.

886

You because there's sons born

is there for 67 days

887

and I'm just texting back and forth

with Dave Brailsford

888

because I'm like, trying

to trying to find an outlet there.

889

And like, I think the tour was

or nos during the Giro or something.

890

And we were talking about

I think it was with Froome when Froome,

891

won that that year and so long

892

story short,

the sky was going out as a sponsor

893

and they were looking for somebody

to buy the team.

894

David found somebody buy the team.

895

But it was Ineos with Jim Ratcliffe.

896

And Jim said, hey, I'll,

well buy the team.

897

But he also has to do this

running project.

898

Dave's like, yeah, sure,

but knew nothing about running.

899

So naturally

I was the only person that he knew that

900

was involved in anything like that was

is the exact same project, to be honest.

901

and he's like, Robbie, can you do this?

902

And I was like, absolutely.

903

If we go and, the

my involvement in the 159 challenge,

904

which was under different sponsorship

905

and leadership, then the breaking two

and Nike completely separate,

906

was completely different.

907

I was I was working on aerodynamics

908

in the weather for braking 2 in 2017.

909

I had to work on all the projects in 2019.

910

So I got to design the course,

got to choose the location,

911

got to put weather stations up

that were tracking the weather

912

so that we could have like really hyper

forecasting neural networks

913

that were, that are going to tell us

when we should run the race.

914

we had a, company

that we used for CFD analysis,

915

and we were running hundreds of hours

of CFD to try to find the aero formation

916

that we came up with that

I, alluded to earlier

917

and there was, one part of it

918

that not a lot of people

know is the, marathon course

919

is, 42.

920

it's it's, a marathon course.

921

It's 26.2 miles. Right.

922

And the, but to measure it,

they use this like, medieval,

923

device that's, called a Jones [Counter]

I call a medieval device.

924

But it's like this, this old device

that, sits on a bike and a, a course

925

measure will come and it takes over,

like, mechanically, to to measure this.

926

Yeah, I've seen them.

927

I have see them. Yeah. I have no idea.

What do you mean?

928

I have one in my garage right now

because I have the bike that we use

929

to measure this distance

and ratify the course

930

when I saw that and I was like, whoa,

wait a minute, here I go.

931

So what happens when you calibrate

this thing and it's off?

932

And it turns out that they know

that the tolerance of this thing is.

933

And the error this thing is, is 1m/km.

934

So and 42.195km,

935

you can have plus or -42m

when you measure the distance.

936

So what they do

is they add 42m on to the course.

937

But if you had already over

measured by 42, now you're up to 84m.

938

They could be wrong.

939

Not likely,

but it could be 12 of course, by 84m.

940

And when you're talking about 26 seconds

that we missed it last time,

941

I'm like, whoa, we need to find out

a better way to do this.

942

So most of my time, I spent 80% of my time

in Vienna that year,

943

and most of my time was

944

we were remapping the course

and we were using we, used lighter

945

to create a 3D mesh of,

the course that we wanted to use.

946

And then we had to find a way

947

to use differential GPS

to map that physically onto the road.

948

And then we would continue to learn

things.

949

And I had this, crew

that would go out with me,

950

that would remap it to the road,

every time that we would make

951

an update to it.

And literally they were like,

952

come on again,

because they didn't really understand,

953

why it was so important.

954

And I would sit over top of them

and look and, and make sure that they're

955

getting these differential GPS poles

exactly where they needed to be.

956

And they didn't.

They spoke German and I spoke English.

957

And like I thought they were my friends,

I don't think they really liked me.

958

But, at the end of the day,

we were able to get the course

959

to be 42.195km plus or -ten centimeters.

960

And so we were able to get spot on.

961

Quick.

962

A quick question, did you actually,

for the race, draw a line

963

where we had to run to make sure he takes

the precise shortest way? Yes.

964

It was somewhat of a mistake

965

because we created

what was kind of like a goat path for him,

966

and turns out if he had stepped

967

one foot on the line, he was disqualified.

968

And because of

969

that, we optimized

and he was doing circuits and,

970

we had optimized the course.

971

So, so perfect

that now that because he was worried

972

about putting a foot on the line,

he wasn't running exactly on it.

973

So he ran more than the, the course

distance because around the curves,

974

he was like off by a meter from what

the optimal line was.

975

And that added distance around,

so there are a lot of things

976

that like on the day you're like, wow,

it's not how I thought it was going to go.

977

But, we were able to,

we were able to get him on for two hours.

978

He did it in 1:59:40.2.

979

So what you're saying is,

and I think our viewers and our listeners

980

can deduct this themselves,

that you're actually

981

the reason why, Mr.

982

Kipchoge

broke this record in the first place.

983

But like you said, using lead,

our technology, you know, all the time

984

and effort that you put into this

on a specific circuit.

985

you know, back in 2023, Kelvin Kiptum

from Kenya,

986

ran the Chicago Marathon,

and he did it in

987

two hours and 35 seconds on, you know,

without all the bells and whistles.

988

That or the support that you gave the,

989

the of 159 challenge.

990

Right.

991

Are we going to see a marathoner

in the near future?

992

Go under two hours on a normal course?

993

Yeah, I hope so.

994

I hope,

I hope there are things that we learned,

995

that that helped that along, too,

because that was the whole point of it.

996

You know, we didn't just do it so that

we could be the first ones to do it.

997

We did it

so that we could move the sport forward.

998

And, it's just like when you saw

somebody break the four minute mile

999

than it all,

you know, more people started to do it.

00:54:51,788 --> 00:54:55,166

and hopefully there are some learnings

that were applied when,

00:54:55,166 --> 00:54:57,127

when he ran in Chicago.

00:54:57,127 --> 00:55:00,130

and, yeah,

00:55:00,255 --> 00:55:03,216

I, I'm sure that we will see it.

00:55:03,216 --> 00:55:04,134

it's just a matter of when.

00:55:06,219 --> 00:55:06,886

Okay.

00:55:06,886 --> 00:55:08,513

why?

00:55:08,513 --> 00:55:10,140

My last question.

00:55:10,140 --> 00:55:12,309

Short answer. You might just laugh.

00:55:12,309 --> 00:55:17,105

I am 52 years old now, and I was a

pretty good solid rider back in my days.

00:55:17,439 --> 00:55:22,277

If I have a comeback tomorrow,

can you make me top ten next year?

00:55:22,277 --> 00:55:23,528

Tour de France. Not.

00:55:23,528 --> 00:55:27,699

Not in ten days,

but next year, 365 days from now.

00:55:27,907 --> 00:55:30,660

Can you add me working together every day?

00:55:30,660 --> 00:55:33,079

Can you make me top

ten in the tour de France next year?

00:55:33,079 --> 00:55:34,581

Well, what do you say to that?

00:55:34,581 --> 00:55:37,584

Do you still have the ability to say, shut

up, legs?

00:55:37,625 --> 00:55:39,711

Hahaha, that's a good answer.

00:55:39,711 --> 00:55:42,005

I guess I can dig that somehow.

00:55:42,005 --> 00:55:42,881

I can dig it still.

00:55:42,881 --> 00:55:45,884

Oh yes my friend, that we can do it.

00:55:46,176 --> 00:55:47,344

Awesome.

00:55:47,344 --> 00:55:49,471

Already did. Was I say Jenzie?

00:55:49,471 --> 00:55:52,724

Now, I think we've figured out your secret

00:55:52,724 --> 00:55:55,727

for being such a great cyclist.

00:55:55,727 --> 00:55:59,439

Was you were always in the breakaway,

so you were always cooling yourself.

00:56:00,357 --> 00:56:01,608

Yeah, right.

00:56:01,608 --> 00:56:02,984

You know. So, yeah.

00:56:02,984 --> 00:56:06,321

Put that into your your training regimen

for for all the ends.

00:56:06,321 --> 00:56:09,532

Robbie because, yeah,

I don't know many people

00:56:09,532 --> 00:56:13,036

that, had is more breakaway

kilometers than all Jensie here.

00:56:13,036 --> 00:56:15,580

But now that now we know

that he was just trying to stay cool,

00:56:15,580 --> 00:56:16,873

he wasn't trying to win races.

00:56:16,873 --> 00:56:19,000

He knew what he's

doing to stay cool. Yeah.

00:56:20,627 --> 00:56:21,336

Hey, Robbie.

00:56:21,336 --> 00:56:23,254

Thanks a million for being our guest.

00:56:23,254 --> 00:56:25,131

It was super interesting.

00:56:25,131 --> 00:56:28,718

I actually did learn a lot of new things.

00:56:28,718 --> 00:56:31,596

I hope I understand

most of it, to be honest.

00:56:31,596 --> 00:56:34,933

But it was fantastic

to have you as our guest, Robbie.

00:56:34,933 --> 00:56:37,519

And hopefully

with some more future project,

00:56:37,519 --> 00:56:39,521

we can have you back as our guest.

00:56:39,521 --> 00:56:43,149

Maybe next year, same time

and we check on you.

00:56:43,441 --> 00:56:46,236

What are you up to in the years to come?

00:56:46,236 --> 00:56:48,279

Thanks a million for being our guest.

00:56:48,279 --> 00:56:49,948

Hey, thanks so much for having me, guys.

00:56:49,948 --> 00:56:52,367

This was fun.

00:56:52,367 --> 00:56:52,450