Estimated fuel and net

DashPilot can estimate what a completed shift's recorded mileage consumed in fuel, and what was left of the shift's recorded earnings once that estimate is taken off. Both figures are estimates, they are built from two assumptions the driver enters, and the app is careful to say so everywhere they appear.

Every amount, figure and route in this documentation is synthetic.

The calculation

estimated gallons   = recorded miles / miles per gallon
estimated fuel cost = estimated gallons * gas price per gallon

and then, on the same screen:

estimated net after fuel       = recorded shift earnings - estimated fuel cost
estimated net per working hour = estimated net after fuel / working hours

The middle step of the first pair is the whole point. Multiplying recorded miles by a price per gallon prices a mile as though a mile were a gallon, and is wrong by whatever the vehicle's fuel economy is. DashPilot works out the gallons first, in one place, and every surface that shows an estimate comes through it.

What is recorded, and what is assumed

Input	Where it comes from
Recorded mileage	Measured from the shift's retained route, exactly as everywhere else in the app
Miles per gallon	A figure the driver enters, recorded on the shift
Gas price per gallon	A figure the driver enters, recorded on the shift
Shift earnings	The amount the driver recorded for the shift

Three of those four are recorded facts. The vehicle's fuel economy and the price of a gallon are assumptions: DashPilot observes no vehicle, reads no pump and has no network access at all.

An assumption is recorded with the shift it was used for

This is the substantive decision behind the whole feature.

Each completed shift keeps its own copy of the fuel economy and the gas price its estimate was worked out under. A driver who changes vehicle in March has not changed what February's shifts consumed, and a fill-up next month at a higher price did not make last month's driving more expensive. If those two figures lived in one global place, every shift ever worked would be silently re-costed the moment either changed, and a history screen would show different numbers today than it showed yesterday.

So:

• Entering different figures later leaves every earlier shift exactly where it is.
• A shift's estimate can always be reproduced from what that shift itself records.

The convenience is separate from the truth, and it is what Settings is for. A shift takes its copy when it starts, from the vehicle the driver selected and the gas price they last entered. From that moment the shift owns its copy: editing the vehicle, deleting it, selecting another or changing the price all change what the next shift records and nothing that has already been recorded.

When the fuel editor opens on a shift that recorded nothing, it fills its fields from those current defaults, or failing those from the most recent shift that did record a pair, and says so. A filled field is a suggestion: nothing is recorded against the shift until the driver taps Save, and no earlier shift changes because a later one, or a setting, records something else.

A shift worked before the defaults existed is not filled in for the driver. It keeps recording nothing until they open its fuel editor and tap Use Current Defaults, which fills the fields from Settings for them to save. Writing today's figures into last month's work would put an assumption the driver never made into their history.

The one correction to a shift still being worked

A shift's assumptions are otherwise editable only once it has finished, and that rule has one deliberate exception. While the running shift's route has recorded no distance, the driver can correct its snapshot from the running shift's own panel: a driver who started in the wrong vehicle should not have to work a whole shift under a figure they can already see is wrong.

It is narrow on purpose:

• Only before any driving is recorded. Once the route has measured a distance, changing what a mile is assumed to cost would restate what those miles are estimated to have consumed, with nothing to point at, so the correction closes. The rule is measured distance and never elapsed time.
• No route evidence is touched. Nothing is deleted or rewritten to make a correction possible.
• It copies, it does not link. The vehicle's current name and economy, and the current gas price, are copied as facts. Settings is not written and a later change to it does not follow.
• It is not a default being reapplied. The snapshot taken when a shift starts may only ever fill an empty pair and happens without anybody asking; this replaces a recorded pair and only ever happens because the driver saved.

Correcting the assumptions of shifts already recorded, as a group, is not a feature DashPilot has. The remedy for one finished shift is its own fuel editor. See Correcting the vehicle.

Which vehicle a shift says it was worked in

A shift that took its economy from a vehicle also records what that vehicle was called, and shows it under the two figures. It is a label and never an input: no estimate, rate or total reads it, and a shift with an economy but no name produces exactly the same figures.

It is a copy rather than a link, so a shift stays intelligible when the profile behind it is renamed or deleted. The name is recorded beside an economy only when it is that vehicle's economy: a driver who types a different figure by hand records no vehicle, because DashPilot does not know which vehicle covers that many miles on a gallon.

The running shift shows the same two facts on its own panel, so a driver with two vehicles who forgot to switch can see it without opening Settings. See Which vehicle the shift is using.

Missing is not zero

A shift with no fuel economy recorded is not a shift that used no fuel, and one with no gas price recorded is not one whose fuel was free. Each absence produces a sentence naming the half that is missing, never $0.00 and never a dash. The same holds for the route: a shift that retained no usable position has no recorded mileage to estimate over, which is a different statement from a shift that recorded no distance.

Two things are deliberately not treated as missing:

• A recorded gas price of zero is a fact. It means the fuel was recorded as having cost nothing, and it produces an estimate of $0.00.
• A route that was measured and covered no distance is a measurement. It consumed no fuel, and the estimate says so.

A fuel economy of zero is refused outright, because it is the divisor and there is no truthful reading of a vehicle that covers nothing on a gallon.

It follows the recorded mileage, including when that changes

The estimate is derived, never stored. It divides the same recorded mileage every other figure on the screen divides, so anything that changes what the route measures changes the estimate with it. Correcting a shift's end time is the case that matters: moving the end earlier deletes the route recorded after it, the distance is measured again from the positions that remain, and the estimate follows, with both assumptions untouched. See Correcting a shift's end time.

Recorded mileage is a floor. Capture can be interrupted, and the distance across a gap is left out rather than guessed, so where the route is partial the screen says what that means in both directions:

• the estimated fuel is a floor: more miles were driven than were recorded, so more fuel was used than this estimates;
• the estimated net is therefore a ceiling: less was left than is shown.

See Recorded mileage.

An estimate is not a recorded expense

DashPilot records operating expenses, including a fuel category, and those records are a completely separate fact. See Recorded expenses.

• Nothing on this page creates, changes or reads an expense, and recording fuel assumptions inserts no expense row.
• No expense total anywhere has an estimate added to it, and no estimate has a recorded cost taken off it.
• Recorded expenses are not part of a shift's estimated net, because an expense belongs to a date rather than to a shift. Net after recorded expenses remains a period figure: see Period summaries.

The overlap the app states rather than resolves

A driver who records the fill-up that paid for these miles now has two figures that describe overlapping money, in two places: a recorded fuel expense dated to a day, and an estimated fuel cost derived for a shift.

DashPilot does not reconcile them, and this is deliberate. It does not know which shifts a tank of fuel was burned on, so any automatic matching would be an attribution the driver never made, and no later screen or export could tell it apart from one they did. That is the same refusal behind an expense having no shift in the first place. The two figures are kept in separate sections, are never added together and are never netted against each other, and the screen says so.

What the screen shows

On a completed shift's detail, under the recorded figures and the gross rates:

Estimated Fuel

• Estimated fuel cost, with Based on recorded mileage under it, or a sentence naming the half that is missing
• Estimated gallons
• Miles per gallon and Gas price per gallon, as recorded for this shift
• Add Fuel Assumptions / Edit Fuel Assumptions

Estimated Net

• Recorded earnings
• Estimated fuel cost, as the amount being subtracted
• Estimated net after fuel
• Estimated net per working hour

The sections sit after everything recorded, deliberately. What the driver recorded and what the route measured are the trustworthy part of the screen, and reading down it should go from the recorded to the estimated rather than mix them.

A shift with no estimate is still fully readable. Every recorded amount, duration, count and gross rate is exactly where it was, and only the estimated sections say they are unavailable and why. A fuel estimate is never a precondition for reading a completed shift.

Working time is the app's existing definition

Estimated net per working hour divides by working duration: the shift's elapsed time less the stretches the driver paused it. That is the same denominator the shift's gross hourly rate uses and the same division, so the two hourly figures on one screen cannot disagree. See Earnings and metrics.

A shift with no measurable working time has no hourly figure, and is told so rather than shown a zero.

The estimated net may be negative

A shift whose estimated fuel came to more than it recorded being paid produces a negative estimated net, and it is shown as one. That is a real outcome of a bad shift over a long drive, not an error to clamp to zero.

Accessibility

Each figure speaks what it is and what it rests on, rather than relying on the caption beside it:

• the fuel cost speaks based on recorded mileage, and the partial-route sentence where it applies
• the gallons speak the unit as a word, because gal reads well and hears badly
• each assumption speaks what it is assumed to be, or that it was not recorded
• the net speaks estimated net after fuel, and the hourly figure names its denominator

Privacy

Nothing here leaves the device, because nothing in DashPilot does. There is no gas-price lookup, no station search, no vehicle service and no network access at all.

The log records that a pair was recorded, changed, removed or refused, and never a figure: not the miles per gallon, not the gas price, not the estimate and not the net. A gas price is also a statement about where and when somebody fills up, which is the kind of value this project keeps out of its logs everywhere. See Privacy.

In an export

A shift's JSON record carries fuelMilesPerGallon and fuelGasPricePerGallon, each an explicit null where the driver recorded none. The estimate itself is deliberately not exported: the file carries the recorded facts and the assumptions, and a reader who wants the estimate applies the rule at the top of this page to the recorded mileage already in the same record. The export format version is unchanged. See History export.

Over a period

A day, a week, a month or a chosen range states what its shifts are estimated to have spent on fuel, and what they are estimated to have been left with after it. Both are on the period summary, under every recorded figure, and both come from the same per-shift arithmetic above rather than from a second one.

Partial coverage is the whole problem

A week can hold six shifts of which four recorded a fuel economy, a gas price and a measurable route. Adding those four estimates gives a real figure. Presenting it as what this week cost in fuel is a claim about six shifts made from four, and that is what the coverage exists to prevent.

So the estimate never travels without the counts behind it, in two units, because either alone can mislead:

Estimated fuel                     $24.81
4 of 6 shifts · 142.3 of 188.9 recorded miles · 75% of recorded miles

• Shifts says how much of the week's work is behind the figure.
• Recorded miles says how much of its driving is. Two short shifts missing their assumptions cost less coverage than one long one, and only the mileage says so.

A percentage is derived from the mileage, because fuel is consumed over distance, and it is never the only thing said: a percentage cannot say which shifts are missing or how long they were. Where every shift is covered the screen says so in as many words, because "no coverage stated" and "complete coverage" must not look the same. Where none is covered it says the estimate is unavailable, never $0.00.

The estimated net is worked out over the shifts that record both halves

The tempting arithmetic is period earnings - period estimated fuel, and it is wrong whenever the two have different coverage: it takes a fuel figure from four shifts off an earnings figure from six and labels the result as the period's.

So the subtraction is performed over the paired subset — the shifts that recorded an amount and have an estimate — which is the rule every rate in DashPilot already follows. Both halves are shown under the figure, so a reader can see what it was worked out from, and when the subset is not the whole period the screen says so:

Estimated net after fuel           $92.00
$100.00 recorded less $8.00 estimated fuel
Worked out over the 1 of 2 shifts that record both an amount and enough to estimate their fuel.
It is not this period's earnings less this period's fuel.

It may be negative, which is a real outcome. Where a contributing route is partial the fuel is a floor and the net is therefore a ceiling, and the screen says that too.

Estimated fuel and recorded expenses are never combined

This is the screen where the overlap documented above becomes visible for the first time: a recorded fuel expense and an estimated fuel cost can appear a few rows apart, and they may describe the same money.

DashPilot keeps them in two sections, each labelled for what it is, and adds them nowhere:

• Net after recorded expenses is recorded gross earnings less the costs the driver entered.
• Estimated net after fuel is recorded earnings less an estimate, over the paired subset.

Neither has the other taken off it, and no figure anywhere subtracts both. Subtracting both under one label would count the same fuel twice, and DashPilot does not know which shifts a tank was burned on. The reconciliation rule does not exist, so the app shows two honest figures rather than one invented one.

Two periods are not compared on an estimate

The period comparison states no estimated figure. Two periods whose fuel coverage differs are not like with like, and two estimated under different assumptions — a different vehicle, a different price — are answers to two different questions. Nothing ranks periods, scores them or declares one more profitable than another.

In History

Each week under View Older Weeks carries the same two figures above its shifts, with the same coverage, and only when the week has them. A week where no shift recorded assumptions simply does not carry the lines, because a summary above a list is not the place to explain an absent estimate: the shift's own detail screen is, because that is where it can be acted on.

What these figures are not

• Not a recorded expense, and not proof that any fuel was bought.
• Not proof of fuel consumed, or of what this vehicle costs to operate.
• Not profit, net income or take-home pay. Nothing for wear, insurance, maintenance, depreciation, phone costs or tax is subtracted anywhere in DashPilot.
• Not a tax figure. There is no deduction, no mileage allowance and no classification of anything as claimable.
• Not a measurement of the vehicle. The fuel economy is whatever the driver typed, and the app neither checks it nor learns it.
• Not a total for the driving that was done, where the route is partial. It covers the miles that were recorded, and says so.