The Reports That Answer 'Are We On Time?' in EDGEBIC

"Are we on time?" is three different questions, and EDGEBIC by User Solutions answers them with three different reports: Late Jobs says which jobs miss their dates and by how much, On-Time Delivery says what percentage actually shipped on time and which unfinished jobs are heading the same way, and Earned Value says whether a job is behind because it is slow or because it was under-planned. Reading one and calling it delivery performance is how plants end up surprised. This deep dive works the arithmetic of all three with real numbers, and shows exactly where each one draws its line.

These sit in the Executive and Supervisor families of the report catalog. Everything below is computed at the moment you open the report, over your current plan and your logged actuals.

Report One: Late Jobs, the Ranked Problem List

Late Jobs is a snapshot with no date range. It looks at every scheduled job that has a due date, finds the latest scheduled end across all of that job's operations, and computes:

scheduled end = the maximum scheduled end date across the job's operations
days late     = scheduled end minus due date, in whole days

A job appears only when days late is greater than zero, and rows sort worst first. Two consequences follow immediately. A job with no due date can never appear, because it cannot be late against nothing. And lateness here is structural: it is the plan missing the date, visible today, not a job that has already blown its deadline.

Each row then gets a status derived from actuals:

• Not started: no operation has an actual start date.
• In progress: at least one operation started without finishing.
• Behind: everything else.

And a blocker hint, derived from percent complete and days late:

Worked example: three jobs, one report

Today is 20 June. Three scheduled jobs:

J-203 never appears: its plan beats its due date by two days. J-201 and J-202 both tie at seven days late, but they are completely different problems. J-201 is in progress at 45% and under 50% with more than five days late, so it reads "Significantly behind plan": the work is moving, just not fast enough. J-202 is at 0% with no actuals, so it reads "No actuals, never started": nothing is wrong with the pace, because nothing has started. One needs capacity, the other needs a supervisor to ask why nobody picked it up. The blocker hint is a starting point rather than a diagnosis, but it routes the two rows to different conversations in a second.

Report Two: On-Time Delivery, the Scorecard and the Warning

On-Time Delivery takes a date range and splits every job by whether it is finished:

effective end = actual end (if every operation is complete)
                otherwise the latest scheduled end
days vs due   = effective end minus due date, in whole days

Then the status:

The headline percentage on the subtitle bar counts completed jobs only: on-time completions divided by total completions. That is the right definition for a scorecard, because a job that has not shipped has not yet delivered anything. But it means the report has two halves, and reading only the headline hides the second one. The Forecast late rows are the jobs that will damage next month's percentage, visible now, while there is still time to act.

Two operational cautions. The report skips jobs with no due date entirely, exactly as Late Jobs does. And the header totals are computed before any grid filtering you apply, so if you group by customer to get a per-customer figure, use the export rather than the header. For the generic definition of the metric behind the report, our on-time delivery KPI guide is the concept reference.

Report Three: Earned Value, Why a Job Is Behind

Late Jobs and On-Time Delivery both answer "is the date at risk?". Earned Value answers "and why?", which is the question that changes what you do about it. It is denominated in hours rather than currency by default, which suits a shop far better than dollars nobody trusts.

Five quantities per job:

From those come the four numbers people actually read:

SPI = EV / PV      schedule performance: below 1.0 means behind
CPI = EV / AC      cost performance: below 1.0 means burning more hours than earned
EAC = AC + (BAC - EV) / CPI     forecast final hours
VAC = BAC - EAC                 forecast overrun, negative means over

Rows sort by SPI ascending, so the jobs in trouble sit at the top of the grid without you sorting anything.

Worked example: a job that is early and over budget

A five-step job, planned and actual hours:

The cut-off falls halfway through the deburr step. Work it through:

BAC = 8 + 12 + 24 + 4 + 2               = 50 h
PV  = 8 + 12 + 24 + 2 (half of deburr)  = 46 h
AC  = 10 + 11 + 30 + 4                  = 55 h
% done = min(100, 55/50 x 100)          = 100%   (capped)
EV  = 50 x 100/100                      = 50 h
SPI = 50 / 46                           = 1.09
CPI = 50 / 55                           = 0.91
EAC = 55 + (50 - 50) / 0.91             = 55 h
VAC = 50 - 55                           = -5 h

The reading: physically ahead, financially behind. Four of five steps are done against a plan that expected 46 hours of progress, so SPI is above 1. But those four steps consumed 55 hours against a 50-hour budget, so CPI is 0.91 and the job will land about five hours over. Nobody needs to expedite this job. Somebody needs to look at why the machining step took 30 hours against a 24-hour standard, because that standard is now quoting every future job of this type six hours light.

That capped percentage matters. A job running 110 hours against a 100-hour budget shows 100% done, not 110%. The overspend is not hidden; it moves into CPI and EAC where it belongs.

The Three Traps in These Numbers

SPI of exactly 1.00 on a job that has not started. When planned value is zero, the index defaults to 1.0 to avoid dividing by zero. It reads like perfection and means "nothing to compare". Always glance at percent done alongside it.

Currency columns that vanish. Earned Value shows BAC, AC, EV, and EAC in dollars only when at least one work center on the job carries an hourly rate above zero. With every rate at the default of zero, the report hides those columns rather than printing a wall of zeroes. Set rates on your work centers and the columns appear on their own.

Job hours versus the sum of station hours. If a routing runs parallel work centers, the same wall-clock hours are logged on every parallel machine. With the primary-hours-only policy enabled, Earned Value, Late Jobs, and Job Progress count the primary path so job totals stay honest, while per-work-center rows and cost stay full-effort. The two can legitimately disagree, and knowing that in advance prevents an hour of reconciliation.

Which One to Open, and When

The sequence that resolves most delivery escalations runs Late Jobs first (is it real?), Work Center Progress second (which step?), and Earned Value third (slow or under-planned?). If the answer turns out to be capacity rather than the job, the next stop is work center overload causes and fixes, and if it turns out to be data, report reading mistakes covers the ways a correct number produces a wrong conclusion.

Every column in all three reports carries its own definition, formula, unit, and worked example values inside the product, one click away on the Column Details button. That is the subject of why every column carries its own definition, and it is the reason nobody in the plant has to remember what VAC stands for.

For what these metrics look like once a shop acts on them, see the documented results; for the wider system, the complete EDGEBIC guide. Want your own late list, computed against your real capacity? Bring your open orders to a demo.