A manager opens the monthly report. The delivery round to six customers cost more than last month. The carrier says traffic. The driver says the customers kept him waiting. The warehouse says loading was fine. Everyone has a story and nobody has a number to put next to it.
The number that is missing is what the round should take. Not what it took last month, which may have been a bad month, and not a rough guess from the map. A figure built from the parts of the round, so that when the real time comes in you can see which part ran over.
This is called a standard time. This article shows how to build one, how to price it by the minute, and how to read the gap when the real month comes in.
Three parts, and one more at each end
The textbook builds a standard delivery journey from three parts. Add two small ones and you have the whole round.
- Loading at the depot. A fixed time (gate, papers, bringing the truck to the dock) plus a variable time that grows with the number of pallets.
- Running. Distance on each type of road, divided by a speed that suits that road type.
- The drops. A fixed time at every stop that does not care how much is delivered (park, find the person, sign), plus a variable time that does.
- The driver's meal and rest.
- Close-out at the depot. Papers, returns, parking.
The fixed and variable split matters most at the drops. A stop with two pallets and a stop with twenty do not take the same time, but both carry the same fixed cost of arriving. The textbook goes one step further: it sorts customers into groups by access and equipment, so a site with a forklift and a dock sits in a faster group than a site where the driver carries by hand. In its UK example the fixed time runs from 10 to 25 minutes depending on the group. Your Thai numbers will be different, and they should come from your own records.
Running time: speeds by road type, and the ceilings
For each leg, use a speed for that type of road, not one speed for the whole trip. Urban roads, main roads outside towns and divided highways are three different speeds.
The law sets the ceiling, not the planning speed. Read in Thai, the Ministerial Regulation on speed rates gives a truck whose vehicle weight is over 2,200 kilogrammes these ceilings:
| Where | Ceiling |
|---|---|
| Bangkok, Pattaya City, a city or town municipality, or a community zone | 60 km/h |
| Outside those | 80 km/h |
| Outside those, with two or more lanes each way and a physical median | 90 km/h |
| A parallel road, wherever it is | 60 km/h |
| Expressway, at ground level | 90 km/h |
| Expressway, above or below ground | 80 km/h |
A posted sign that sets a lower limit overrides all of these.
Two uses for the table. First, a sanity check: if your standard implies a speed above the ceiling for that road type, the standard is wrong, whatever the map says. Second, a reminder that the ceiling is the fastest a standard could ever be. The speed you plan with has to come from your own records, and the way to get it is in how long inland transport really takes.
One more line in the same regulation is easy to miss. A vehicle while towing another vehicle has lower ceilings in every row: 45, 55 and 65 km/h in place of 60, 80 and 90. The regulation read here does not say whether a tractor pulling a semi-trailer counts as towing. Ask the carrier, and if the answer is yes, use the lower row for those trucks.
The law also limits continuous driving to four hours, after which a continuous half hour of rest allows up to four more (Land Transport Act Section 103 bis). A long round needs the rest built in. The standard should contain the half hour, not discover it on the day.
A worked round
All numbers below are an illustration. Replace each with your own.
A six-wheel truck loads 12 pallets at a depot on the edge of Bangkok and delivers to six customers.
| Part | How the standard is built | Standard |
|---|---|---|
| Loading | 30 min fixed + 12 pallets Γ 0.5 min | 36 min |
| Running, urban | 45 km at 25 km/h | 108 min |
| Running, main road with median | 60 km at 60 km/h | 60 min |
| Running, other road | 30 km at 40 km/h | 45 min |
| Drops | Fixed 15, 15, 15, 25, 25 and 40 min by site, + 12 pallets Γ 3 min | 135 + 36 = 171 min |
| Meal and rest | Allowance | 30 min |
| Close-out | Allowance | 15 min |
| Round | 465 min (7 h 45 min) |
Two things fall out before any real data arrives.
How full the day is. If the truck is hired for a 10-hour day, 600 minutes, the round fills 465 of them, or 77.5%. That leaves 135 minutes of room. It is the room you have to absorb a late customer or to add a seventh drop.
A price per minute. Suppose the vehicle-day costs 3,600 baht (driver, truck, fuel allowance, all in, an illustrative figure). Over 600 minutes that is 6 baht a minute. The standard round is 465 Γ 6 = 2,790 baht of the day. This is a price tag for comparison, not a bill. On a truck hired by the day, minutes inside the paid day cost no extra cash, but they use up the room. Minutes beyond the day cost real money. How the day is built as a denominator is in choosing the unit you measure transport in.
Reading the month
Now the month comes in. The average actual round, taken from gate times and tracker records over 22 working days, is 533 minutes. That is 68 minutes over standard. Here is the same table with the actuals beside it.
| Part | Standard | Actual | Over |
|---|---|---|---|
| Loading | 36 | 58 | +22 |
| Running | 213 | 223 | +10 |
| Drops | 171 | 207 | +36 |
| Meal and rest | 30 | 30 | 0 |
| Close-out | 15 | 15 | 0 |
| Round | 465 | 533 | +68 |
Without the standard, the report says "the round took 533 minutes" and the argument starts. With it, three things are visible at once.
- Running is nearly innocent. It is 10 minutes over, 15% of the gap. The carrier's "traffic" explanation explains the smallest piece.
- Loading is your own dock. 22 minutes over, 32% of the gap, on the part you control. That one is yours to fix, not a reason to argue with the carrier.
- Drops are over by 36 minutes, but look one level down. Say the six sites are logged separately. Five of them are within six minutes of standard in total. The sixth, the site with the 40-minute fixed time, was set at 46 minutes (40 + two pallets Γ 3) and took 76. That one site is +30 of the 36.
One site out of six explains 44% of the whole overrun (30 of 68). It is not a driver problem, not a carrier problem and not a traffic problem. It is a customer whose gate or dock does not work the way the standard assumed. The options are to renegotiate their window, to change how the truck is received there, or, as in multi-drop or two trucks, to put that site on its own truck. Each is a conversation with a name on it, which is more useful than a general complaint about delivery performance.
The cost of the overrun, at the same 6 baht a minute: 68 Γ 6 = 408 baht per round, about 9,000 baht over 22 rounds. Small in baht. What matters more is the room. The day had 135 minutes spare in the standard and now has 67 (600 minus 533). One more bad stop and the round runs past ten hours.
Time is one question, price is another
The textbook names three causes of a cost deviation: the amount of work changed, the efficiency changed, or the price changed. A standard time answers the efficiency question on its own, because it measures minutes, not baht.
Suppose the vehicle-day goes from 3,600 to 3,900 baht, so a minute goes from 6.00 to 6.50. The time variance is unchanged at 68 minutes. The price effect is separate: 533 minutes Γ 0.50 baht = about 267 baht a round. Kept apart, you can tell "the round got longer" from "the minute got dearer", and you can test a rate increase request against the right half. How to test one is in testing a rate increase request.
Where a standard goes wrong
- A stopwatch ride-along gives a flattering number. The textbook warns that time-study data is expensive and hard to trust because people under observation do not behave naturally. Build from records (gate-in and gate-out times, tracker stop times, signed delivery times) over eight to twelve weeks, not from one observed day.
- An average hides the bad site. Keep the standard per site for drops and per road type for running, as above. The one-line average round time is the version that hides the one site.
- A standard nobody owns is ignored. The textbook's good-practice list includes user ownership: the people who run the round should help build the standard. Show the dispatcher and the driver the table before it becomes a target.
- A standard that costs more than it saves. Do not build one for every route. The rounds that run most often, and the ones that go over most often, come first.
- Standards go out of date. A new customer, a new road layout or a changed dock changes the parts. Revisit a site's fixed time whenever its access changes.
- A standard is not a whip. Used to punish drivers for minutes they cannot control, it teaches them to hide the cause. How the pay scheme changes this is in how driver pay changes how drivers drive.
What to do on Monday
- Pick the three rounds you run most often.
- For each, list the five parts and write down the time you believe each takes. This is your first standard, rough but explicit.
- Pull eight weeks of gate times and tracker stop times. Replace each guess with the measured figure, using the typical day rather than the best or worst.
- Check each running speed against the ceiling for that road type. If a speed is above it, the standard is wrong.
- Divide the vehicle-day cost by the minutes in the paid day. That is your price per minute.
- Each month, put actual beside standard for the five parts, and for drops, for each site. Talk about the largest gap first, and name the owner.
Where the evidence stops: no Thai study of standard times for delivery rounds was found, so the loading, drop and rest times, the planning speeds, the vehicle-day cost and the whole month of actuals above are an illustration of the method and not Thai measurements. The method comes from one UK textbook. The speed ceilings come from the Thai regulation, read from a single host, and the regulation as read does not settle whether a tractor with a semi-trailer counts as towing. Section 103 bis was read from a PDF whose text layer is scrambled, with the numbers and key words legible.
