What is ready, and what is waiting
0 drives, 0 comparable. 1 of 8 learned quantities have met their gate. A quantity is promoted when the data meets its gate, not when a figure looks convincing.
| Quantity | Status | Value | On what evidence |
|---|---|---|---|
| Best-ever block times Waiting for: 3 crossings, and a best crossing of at least 5 s so a clipped corner cannot set it | no data | 0 | 0 of 0 blocks crossed 3+ times |
| Regeneration limit Waiting for: 200 samples in a speed bin, and speed above 15 km/h where the relation holds | not yet | 0 | 0 speed bins with 200+ samples |
| Where traffic holds you up Waiting for: 3 crossings of every kilometre in the stretch | not yet | 0 | 0 of 0 hold-ups rest on enough crossings |
| When to leave Waiting for: 20 trips covering at least 5 distinct quarter-hour slots | not yet | 0 | 0 comparable trips across 0 quarter-hour slots |
| External traffic estimate | ready | 3,500 | 3500 poll(s) folded, 380 bucket(s) past the sample gate |
| Charging efficiency Waiting for: 5 charges with both sides measured | no data | — | no charge yet has both sides known |
| How much wind the car meets Waiting for: the residual must improve by at least 2 percent across the scanned range | not yet | 100% | 0 windows over 0 drives; residual is flat across the scan |
| Drivetrain loss Waiting for: the residual must vary by at least 5 percent across accessory loads from 0 to 2400 W | not yet | — | trades off against accessory load; report the agreement ratio |
Every kilometre, accounted
The odometer is absolute, so the distance between one log's last reading and the next log's first is exactly what was driven unrecorded. It measures distance and never purpose, so it is never folded into a corridor.
| Where the driving went | km | Drives |
|---|
Not enough to model
A state this tool will sit in for a while, not an error.
The model needs drives long enough to be commutes, and map data along them. Neither is present yet.
What does speed cost?
Steady 30-second stretches, corrected for climb and descent. 0 windows from 0 trips. Shown as kWh/100km until a charge sits between two logged drives with the odometer still, which is what measures this car's own pack capacity rather than a nominal one.
Not enough steady-state driving yet.
By speed bin
| Speed | Windows | Road kW | kWh/100km | Range, km |
|---|---|---|---|---|
| No bins yet | ||||
Time against energy
Needs a long trip with mapped road data to price.
Needs at least two speed bins.
Every channel against distance along the route, drives overlaid, one cursor across all of them. Up to four drives per direction, longest first.
Not enough to trace yet
Needs drives with position data.
Not enough to compare yet
Needs two drives over the same road in the same direction.
The route
Drawn from your own GPS trace and shaded by typical delay. No basemap, no tiles, no network — the shape is the orientation.
No position data yet.
This page now contains your route. It was built with --map, so unlike the default it shows where you drive. Fine on your own machine; think before screenshotting or sending it. Rebuild without the flag for a version safe to share.
Where the time goes
Route split into 250 m direction-aware blocks. 0 of 0 blocks have been driven three or more times, which is what it takes to know what free-flow looks like there.
Needs blocks driven at least three times.
By trip
| Trip | Blocks matched | Time lost | Sustained delay began |
|---|---|---|---|
| No repeat blocks yet | |||
Free-flow for each block is currently the fastest of only a few passes, so these delays are measured against the best you have ever done there. The shape — which stretches lose time — is meaningful now; the magnitudes settle as blocks accumulate passes.
Against the speed limit
Legal minimum is the posted limit with no traffic and every light green, plus a 9 s penalty per stop sign. Speed limits and stop lines from OpenStreetMap.
| Trip | Blocks | Legal minimum | Best vs legal | Traffic |
|---|---|---|---|---|
| Run with --roads to fetch road data | ||||
"Best vs legal" is usually negative on this route — about −4 minutes across a full commute — because your best runs beat the posted limits. That is a statement about how the road is driven, not an error. Where it goes positive, the road is holding you below its own limit no matter how clear the traffic is.
Trips
| Departed | Min | km | km/h | kWh | /100km | °C | Head- wind | Cross- wind | Conditions | Status |
|---|
Measured against the published road load
Not enough steady cruising to compare yet.
What more driving unlocks
- ○Speed / energy shape3+ comparable trips
- ○Efficiency vs temperature10+ trips across a temperature range
- ○Departure-time trend20+ trips spread across departure times
- ○Departure-time model with uncertainty40+ trips
- ○Central store worth building150 trips logged (0 so far) — CLOUD-001 is deferred until then
What this cannot tell you yet
Wind and rain are measured, not yet corrected for. Headwind is the wind along your direction of travel; crosswind is the part across it, which costs energy too but cannot help you. Rain and snow raise rolling resistance and slow traffic, confounding the energy figures and the delay figures at once. Snow is flagged separately from rain because a wintry trip is not a summer trip with a correction applied — it arrives with the cold that makes the cabin heater and pack conditioning expensive, and it belongs in its own group once there are enough of them.
Departure time is not yet analysable. It needs many trips rather than many samples, and no amount of processing substitutes for driving.
Trips marked app energy are excluded from comparisons. Those logs lacked pack current, so energy came from OBD Fusion's own total, which scatters ±17% against integrated voltage × current.
No charging data
A state the tool is designed for, not an error.
No charging has been fetched yet. Charging appears nowhere in the vehicle logs — the capture app is not running while the car is plugged in — so the circuit monitor is the only witness to it.
Run emporia_fetch.py fetch to populate it.