The On-Time Job Count Metric in Optimization

On-time job count is the plainest delivery measure the optimizer tracks: how many jobs in a plan finish by their due date. Its mirror, late job count, is how many finish after, and together they cover every scheduled job. The optimizer computes both with the same shared calculator that produces every other measure, and shows them in the comparison. Where weighted tardiness measures how badly jobs miss, on-time job count is a simple headcount of who made it. Most goals use it as the tiebreaker right after weighted tardiness, because among plans that are equally bad on total lateness, fewer distinct late jobs is usually easier to manage.

EDGEBIC by User Solutions surfaces on-time and late job counts in every optimizer comparison. This post explains how they are computed, how they break ties in the goal ranking, and why you read them alongside weighted tardiness rather than instead of it. For the full set of measures, see how the optimizer scores a schedule, and for the severity measure it complements, minimizing weighted tardiness.

How the count is computed

The calculator checks each job's finish against its due date. A job that finishes on or before its due date is on time; a job that finishes after is late. Sum the two and you get the on-time job count and the late job count, which always add up to the total job count in the plan. There is no partial credit and no weighting: a job one minute late counts the same as a job a week late in this measure, and a job one hour early counts the same as one exactly on time. That plainness is the point. On-time count answers one question cleanly: how many delivery promises did this plan keep.

Why plainness is useful

A headcount is easy to communicate and easy to act on. "This plan lands 27 of 30 jobs on time, the current one lands 24" is a sentence a planner can take to a production meeting without explanation. It maps directly to the on-time delivery percentage most shops report to customers and management. Where weighted tardiness is the right measure for the optimizer to rank by, because it captures severity and priority, on-time count is often the right measure to report, because everyone understands it. The comparison shows both, so you optimize by one and communicate with the other.

How it breaks ties in the ranking

The default On-time first goal ranks weighted tardiness first and late job count second. This ordering matters. Weighted tardiness leads because it is the most complete view of delivery health. But many real schedules have several orderings that tie on weighted tardiness: different plans that are equally late in total weighted hours. Late job count breaks that tie, and it breaks it the way a plant prefers.

Consider two plans with the same weighted tardiness. Plan A spreads the lateness across three jobs, each slightly late. Plan B concentrates it in one job that is badly late. The goal prefers Plan B, fewer distinct late jobs, because one late job is often easier to manage than three: you can expedite it, call one customer, or accept one slip. Three slightly late jobs mean three conversations. The tiebreak encodes that practical judgment. Read how the tiers compare in how the optimizer scores a schedule.

Reading on-time count and weighted tardiness together

The two measures can move independently, and reading them together tells you what kind of improvement a plan made.

That last case is worth a closer look in the move list, since it means the plan shifted which jobs are late rather than simply reducing lateness. For reading the moves, see reading the optimizer move list.

What it does not measure

On-time count deliberately ignores severity and priority. It will not tell you that the three jobs a plan made late are your three biggest customers, or that the one it rescued was minor. That is by design: it is the headcount, and severity is weighted tardiness's job. This is exactly why the goal ranks weighted tardiness first and uses late job count only to break ties. If you ranked on-time count first, the optimizer might flip several trivial jobs to on time while letting a key job slip badly, because the headcount would improve. Ranking severity first and headcount second protects the important jobs and then tidies the count, which is the order a delivery-driven shop wants.

The bottom line

On-time job count is the number of jobs that finish by their due date, a plain headcount that adds up with late job count to cover every job. It is easy to report and it breaks ties in the goal ranking right after weighted tardiness, preferring plans that concentrate lateness in fewer jobs. Read it alongside weighted tardiness: the count tells you how many landed, the weighted measure tells you how big the misses are. Both appear in the comparison after every run. To see them, open the Optimizer tab in Schedule Jobs and run against your current plan. For the wider picture, read the EDGEBIC optimizer guide, and to explore the platform, visit EDGEBIC.