COACH JOE BEER

BIKEROBOT v.11.26

Performance engine for pacing, aerodynamics and race fuelling.
Weather: Open-Meteo when loaded
Race Prediction: load a GPX/FIT course, enter known athlete inputs and let SPEEDFIX predict pacing, bike split, fuel and hydration. Course elevation may be smoothed to reduce GPX noise.

1. Course

Upload a GPX or FIT course
GPX track points or FIT GPS records with elevation work best. Wahoo ELEMNT/BOLT activity FIT files are supported where GPS records are present.
Course Library
Stores GPS/elevation only β€” never athlete power, speed, heart rate or other ride execution data.
No GPX or FIT course loaded.

2. Athlete & Bike

3. Aerodynamics & Road

4. Bike Fuel, Fluid & Sodium

Live race balance trends
These outputs use the last predicted event duration and update instantly as fuelling, fluid, sweat or sodium inputs change. Weather-adjusted sweat rate is used when a weather profile is loaded.
Finish fluid balanceβ€”
Finish sodium balanceβ€”
Finish hydrationβ€”

Sweat loss vs fluid intake

Sodium loss vs sodium intake

Hydration % through the event

Sweat lossβ€”run SPEEDFIX
Fluid intake / balanceβ€”β€”
Sodium intake / balanceβ€”β€”
RQ is calculated from planned power relative to FTP/RQ1. During prolonged events it can drift downward, but SPEEDFIX V34 blunts that drift as carbohydrate availability rises; it never models a zero-carbohydrate race strategy.

Open-Meteo Event Weather V34

For outdoor sessions, SPEEDFIX fetches hourly Open-Meteo data from the course location, event date and start time. V34 automatically uses forecast data for current/future events and Open-Meteo historical weather for past events. The same module also imports Open-Meteo CSV/XLSX downloads like the sample files supplied. Indoor sessions ignore weather.
Load a GPX/FIT course and choose an event date/time, then press Load Open-Meteo Weather.
Weather advisory will update after data are loaded.
Weather attribution: Open-Meteo. Free API use is subject to Open-Meteo's licence/usage terms; production client use may require a commercial plan.

Actual Athlete Ride / Run / Race Import RACE ANALYSIS

Import athlete ride, run or race

FIT files are analysed from the athlete’s recorded power trace without smoothing the power data. Real-world FIT files may also supply GPS/elevation; indoor FIT files from a smart trainer, Wattbike, Zwift or ROUVY can be analysed without a GPX course. ZWO files are workout targets rather than measured execution.

Analyse keeps the athlete's real power trace. Extract GPS Course Only removes power/speed/HR and uses only route coordinates/elevation. Copy Surge Pattern uses the measured hard-effort behaviour in the prediction model.
No athlete ride imported.
Near-threshold segments (β‰₯90% RQ1) and uphill segments already prescribed at β‰₯85% RQ1 are adjusted by this percentage. V34 then recalculates speed, glycogen use and accumulated supra-threshold fatigue. It is designed to show why pushing harder on the difficult parts can add substantial fuel cost for surprisingly little bike-split gain.

Coach Scenario Lab V34

Create a named improvement scenario without changing the athlete’s baseline inputs. The athlete-facing view shows the outcome, not the underlying equipment library or adjustment buttons.

These are editable coaching assumptions for modelling, not universal product claims. Replace them with measured CFD, field-test, tyre, drivetrain or athlete data whenever available.

Saved scenarioMassPowerDragBike timeChangeAction
No saved scenarios yet.

7. Run SPEEDFIX

SPEED FIX 34 uses segment-by-segment gradient, CdA, rolling resistance, gravity, drivetrain loss, pacing rules, substrate use and a descent speed cap. Imported FIT or ZWO data can then be compared with the prescribed model, while the Pacing Accuracy control tests the cost and limited speed return of harder-segment overshooting.
SPEED FIX 34 β€’ performance summary

Run the model.

Upload a GPX/FIT course or load the demo route, then calculate a baseline and optimisation scenario.

Scenario versus baseline
β€”
No baseline comparison yet.

Performance Results

Predicted bike time
β€”
Course total
Average speed
β€”
km/h
Distance
β€”
km
Elevation gain
β€”
metres
Average power
β€”
watts
Normalised power
β€”
watts
Intensity factor
β€”
NP / FTP
Variability index
β€”
NP / average power
Estimated Bike Stress
β€”
IFΒ² Γ— hours Γ— 100
Mechanical work
β€”
kilojoules

Race Analysis β€” Actual Power Execution

Import an athlete FIT file to analyse the recorded power trace. SPEEDFIX does not smooth measured power in this mode.
Ride time
β€”
analysed intervals
Average power
β€”
watts
Normalised power
β€”
watts
Variability index
β€”
NP / average
Power CV
β€”
standard deviation / mean
Zero-power time
β€”
coasting / stopped
>105% threshold
β€”
time above
>110% threshold
β€”
time above
Spikes >105%
β€”
separate excursions
Max power
β€”
watts
30 s peak
β€”
rolling average
5 min peak
β€”
rolling average
Carbs oxidised
β€”
actual exertion demand
Carbs absorbed
β€”
race-plan availability
Finish glycogen
β€”
% of starting store
No actual ride analysed yet.

Peak Power Profile

Apparent CdA vs Power OUTDOOR BIKE

The AiRO/Bioracer CdA input remains the measured reference. V34 also estimates apparent CdA from recorded power, ground speed, gradient, mass, rolling resistance and time-matched Open-Meteo wind/air-density data. This is a diagnostic estimate, not a replacement for CFD or aero testing.
Measured/reference CdA
β€”
Median apparent CdA
β€”
10–90% apparent range
β€”
Usable samples
β€”

Carbohydrate In vs Carbohydrate Out

Carbohydrate in follows the entered race feeding schedule and absorption assumption. Carbohydrate out follows the athlete’s recorded power interval-by-interval. At RQ 1.00 / threshold, SPEEDFIX uses the calibration threshold watts = carbohydrate grams per hour; above threshold, carbohydrate cost rises progressively.
Absorbed carbohydrate in Cumulative carbohydrate out Glycogen remaining Power / threshold
V31 surge sensitivity compares the same course with +5% and +10% effort on slower/harder sections, showing the speed return against extra carbohydrate cost.
Execution checkResultInterpretation
No actual ride analysed yet.

Athlete Ride vs SPEEDFIX Prescription V34

Run SPEEDFIX to see the pacing-behaviour comparison. Import a FIT or ZWO file in Coach View to add athlete data.
Pacing-adjusted time
β€”
hard-segment behaviour
Time change
β€”
versus prescription
Adjusted average power
β€”
watts
Adjusted carbohydrate
β€”
grams oxidised
Adjusted finish glycogen
β€”
% of start
Fatigue penalty peak
β€”
later effective power
Imported source
β€”
FIT / ZWO
Imported time
β€”
moving / workout time
Imported average / NP
β€”
watts
Prescription accuracy
β€”
weighted power match
Hard-effort execution
β€”
% of prescribed
Time above 110%
β€”
RQ1 threshold
PrescribedPacing-adjustedImported ride/workout
VersionTimeAverage powerNPCarbohydrateFinish glycogenFatigue load
No model run yet.

Elevation, Power & Predicted Speed

Race Fuel Summary

Run a calculation to estimate carbohydrate, fluid and sodium outcomes.
Glycogen finishβ€”
Hydration finishβ€”
Sodium balanceβ€”

Granular Glycogen, Carbohydrate Demand & Hydration

Every course or imported-ride interval has its own power, RQ, carbohydrate fraction and glycogen demand. The middle panel exposes the changing carbohydrate burn rate and power relative to RQ1, making climb spikes and low-power descents visible rather than smoothing them into a near-constant decline. Red lines flag glycogen below 25%, hydration below 96%, and power above 110% of RQ1.

1% Better β€” Live Performance Sensitivity

Use the sliders for a live combined scenario. This does not overwrite the named Coach Scenario Lab above.
Run a calculation to activate the sliders.

Common coaching scenarios

ScenarioChangeBike timeTime savedAverage speed
No calculation yet.

Coach Notes

Coach notes remain hidden in Athlete View and are not included in the athlete PDF.

Athlete Action Plan

No athlete action plan has been entered.

Highest-Time Course Segments V34

A quick coaching view of where the athlete spends the most time. It helps prioritise pacing, position durability and fuelling cues; it is not a cornering or technical-course model.
SegmentDistanceGradePowerSpeedTimeRQCarbohydrate
No calculation yet.

Run Athlete & Shoes RUN ONLY

These inputs are used only by the running engine. No bike mass, FTP, CdA or cycling descent settings are used here.
V30 running-energy calibration: level-running energy cost is modelled around 0.95–1.05 kcal/kg/km before terrain and shoe adjustments. Planned pace is taken from the average-pace input; CP, VOβ‚‚max and Z1 pace are used to check whether that pace looks physiologically credible for the predicted duration.

Run Fuel & Hydration RUN ONLY

Run sodium advice compares estimated sweat sodium loss with planned sodium intake. It is advisory rather than a hard performance limit, with larger mismatches becoming more relevant as duration and heat increase.

Run Prediction Plan V30 RUN

Prediction uses running-specific pace and power anchors, athlete mass, shoe economy, terrain, fuelling and pacing strategy.

Run Prediction Results

Planned race time
β€”
from planned average pace
Planned average pace
β€”
min/km
Expected power / CP
β€”
feasibility check
Energy cost
β€”
kcal / mile range
Total energy
β€”
kcal range
Carbohydrate use
β€”
low–high grams
Planned carbohydrate
β€”
grams consumed
VOβ‚‚ demand
β€”
% VOβ‚‚max estimate
Time above Z1 pace
β€”
context, not a hard limit
Feasibility
β€”
duration-aware check
Duration-aware CP guideβ€”
Pace-derived oxygen costβ€”
Pace / power consistencyβ€”
Carbohydrate share bracketβ€”
Run the prediction to generate the athlete pacing model.
Sodium advisory will appear after calculation.

Run Scenario Lab V34

Save different athlete/race assumptions without losing the current reference. Planned pace remains the primary race target; scenario changes test whether a lighter athlete, stronger CP, different shoe economy or pace target materially changes feasibility.
ScenarioPaceTimeCP %Carb rangeFeasibilityAction
No saved run scenarios yet.

Run Analysis β€” Actual Run / Race V30 RUN

Import a running FIT file from Stryd, Garmin, COROS, Suunto, Zwift or another compatible source. Analysis uses running Critical Power and recorded run pace/power only β€” cycling FTP and bike inputs are not used.
No run imported.
Run time
β€”
Distance
β€”
km
Average pace
β€”
min/km
Average power
β€”
watts if recorded
Power / CP
β€”
>105% CP
β€”
Import a run FIT file to assess execution.

Peak Running Power Profile

Important SPEED FIX 34 Limitations

PrototypeSPEED FIX 34 is an engineering and coaching estimate, not a medical or guaranteed performance prediction. It does not yet model cornering and braking zones, traffic, changing road surface, yaw-dependent CdA, directional wind by course bearing, drafting, thermal strain, altitude acclimation or a fully validated Wβ€²/critical-power depletion system. The V34 fatigue penalty is a coaching sensitivity model, not a physiological diagnosis. The in-browser FIT reader targets standard activity record fields, including common Wahoo ELEMNT/BOLT record layouts and may not decode proprietary developer fields. GPX elevation noise and inaccurate CdA, Crr, weather, sweat-rate or RQ inputs can materially change prediction results. Apparent CdA from an actual ride is an inverse-model estimate and is sensitive to wind error, acceleration, gradient, tyre Crr and power-meter accuracy; compare it with AiRO/Bioracer/CFD measurements rather than replacing them. Race Analysis uses the imported recorded power trace without smoothing; its accuracy therefore depends on the recording device, power-meter calibration and FIT sampling quality.