RPD Vaktari Statistic · User manual

01What it is and what it does

Every few minutes the vision system saves a .DATA file with the summary of the lot being made: how many parts so far, how many good, bad and not seen, and the statistics of each characteristic. RPD Vaktari Statistic reads all those files from the folder and puts them in order, so you can see how a lot went, compare it with earlier ones and produce a report.

You look at it with a browser, from the same computer or from any other on the network. It has four tabs:

Tab	What it is for
History	The counters and characteristics of a reference or a lot, snapshot after snapshot.
References	What was made each day: the index for finding a reference or a lot.
Report	The full report of one snapshot, to print or save as PDF.
Export	Getting the data out as CSV, or the report as PDF.

The three counters

Counter	What it counts
OK	good parts
NOK	bad parts. The machine breaks them down into NOK1, NOK2 and NOK3.
NV	parts not seen: they went through the machine but the vision system did not see them, so they could be given neither as good nor as bad

Three words you will see everywhere

Snapshot. Each .DATA file is a snapshot: a picture of the counters and the statistics at that moment. The counters are cumulative: each snapshot includes everything the previous one already said.
Interval. What happened between one snapshot and the previous one. Vaktari works it out by subtraction: if one snapshot says 1,000 parts and the next 1,300, 300 were made in the interval.
Lot. The file carries no lot number. Vaktari takes a lot to be the period between two resets of the machine's counters: when a counter is lower than in the previous snapshot, a new lot begins. That is why lots are named by the date and time they start.

Vaktari recalculates nothing The files only carry summaries — mean, σ, Cp, Cpk and histogram — not the part-by-part measurements. All these values are shown as the vision system wrote them.

02Starting and access

On the computer where it is installed

Double-click vaktari_statistic.bat, or the shortcut that points to it. A black window with text opens: it is the program that reads the files. Do not close it; you can minimise it.
Open the browser and go to http://localhost:8763.

From another computer on the network

Type http://PC-IP:8763 in the browser, where PC-IP is the address of the computer running Vaktari. Whoever did the installation will give it to you.
It is worth saving as a bookmark: it is the only thing you need to remember.

No need to refresh

The program looks at the folder again every 15 seconds and the page updates by itself. When opened, it goes to the most recent lot, which is the one being made now; and when that lot ends and another begins, it jumps to the new one.

03The header

Vaktari Statistic header with numbered callouts — **Fig. 1** · The header and the tabs.

1: Brand and module.
2: Where the data comes from. The name of the folder or line and, below it, the reference being looked at.
3: Language Catalan, Spanish, English, Basque, French and Portuguese. The browser remembers the choice.
4: Theme Dark or Light. Also remembered. Green, red and amber mean the same in both.
5: Tabs. The browser remembers the last one you had open.

Language drop-down — **Fig. 2** · The six languages and the two themes.

Theme drop-down — **Fig. 2** · The six languages and the two themes.

04“History” tab

This is the main tab. Everything on it corresponds to the selection you make in the bar at the top, and that same selection is the one used by Report and Export.

4.1 · The selection

Selection controls of the History tab — **Fig. 3** · What you want to see.

1: From and To The period.
2: Reference One of those with snapshots in the period, or All. The characteristic tables are only shown with a reference chosen.
3: Lot One specific lot, named by its start date and time, or All. Changing the dates or the reference sets it back to All.
4: Shortcuts. Last day is the last day with data; 7 days and 30 days end on that same day; All goes back to the first snapshot in the folder.
5: Back to all lots Only shown when a lot is chosen.

4.2 · The counters of the selection

Machine total The parts given by the machine's total counter and, below, how many lots and snapshots the selection holds.
OK, NOK and NV Good, bad and not-seen parts, with their percentage. Under NOK, the breakdown the machine gives (NOK1, NOK2…).
Worst Cpk Shown when there is a single reference with dimensional characteristics: the characteristic with the lowest Cpk in the last snapshot.
Last snapshot The time and day of the last snapshot in the selection.

If the selection spans more than one lot, the figures are the sum of what was made in each interval, not the counter value.

The two notes below The total does not match. The machine total does not always equal OK + NOK + NV. When that happens, Vaktari says so, with the difference, and computes every percentage over OK + NOK + NV.
Counter resets. If the selection contains any reset, it says how many. On the charts they are marked with a vertical line.

4.3 · Lots

Lots table — **Fig. 5** · One row per lot, the most recent at the top. The highlighted row is the chosen lot.

For each lot: when it starts, the reference, the last snapshot, how many snapshots it has and the counters of its last snapshot, with the NOK percentage. Clicking a row opens that lot: the whole tab then refers to it alone.

4.4 · Cumulative counters and NOK per interval

Charts of cumulative counters and of NOK and NV per interval — **Fig. 6** · Two charts on the same time axis. One point per snapshot.

1: Cumulative counters Total, OK, NOK and NV as each snapshot gives them. In a lot that is going well, the lines rise at a steady rate; a flat stretch is the machine stopped. The dotted vertical line is a counter reset.
2: NOK and NV per interval Of the parts made between one snapshot and the previous one, what percentage was NOK and what NV (not seen). It is the chart that shows when things went wrong: the cumulative figure hides a bad quarter of an hour, the interval does not.

Hover over a point to see the time of the snapshot and the values.

4.5 · Dimensional characteristics

Dimensional characteristics table on the History tab — **Fig. 7** · Every dimensional characteristic of the reference, from all cameras. The highlighted row is the characteristic shown in the trend.

With a reference chosen, the capability table appears here with the values of the last snapshot in the selection; the title says which. The columns are the same as in the report (fig. 17), where they are explained one by one. The status on the right sums up the Cpk: capable, borderline or not capable.

Clicking a row opens the trend of that characteristic just below.

If the reference has no dimensional characteristics, it says This selection holds no dimensional characteristics. With All references, it asks you to choose one.

4.6 · Trend of a characteristic

Trend of a dimensional characteristic: Cpk and mean — **Fig. 8** · The trend of a dimensional characteristic: Cpk on the left and the mean between limits on the right.

The Characteristic drop-down chooses which one, with the camera in front; you can also click a row in the tables. Depending on the kind of characteristic, different charts appear:

Chart	What it shows	Shown for
Cpk	the Cpk of each snapshot, with a green line at the target and an amber one at 1	dimensional characteristics
Mean and limits	the mean of each snapshot between the tolerance limits and the nominal, with the measured minimum and maximum as dotted lines	dimensional characteristics
NOK per interval	how many parts this control rejected between one snapshot and the previous one	attribute characteristics

In the example above, the Cpk stays just under the green target line, and the mean runs below the nominal but far from the limits: the characteristic is stable, slightly off-centre. The dotted vertical line is a counter reset.

Trend of an attribute characteristic — **Fig. 9** · The trend of an attribute characteristic: it only has NOK per interval.

Each point is the value the vision system had computed at that snapshot. With a single snapshot there is no trend yet, and Vaktari says so.

4.7 · Pareto of defects

One row per control, from most to least, with the camera in front. The bar is the number of NOK parts that control counted during the selection, and the column on the right, the cumulative percentage.

4.8 · Attribute characteristics

Attribute characteristics table — **Fig. 11** · The pass / fail controls, and those that only have a counter.

These are controls with no nominal or Cpk. For each one: parts checked (n), not measured, below the minimum, above the maximum, the NOK counter of the last snapshot and the NOK of the whole selection. A dash means the file does not carry that value: there are controls for which the machine only saves the counter. Clicking a row opens its trend.

05“References” tab

It is the index of everything that was made. It ignores the reference and lot chosen in History; it only follows its own period, which on opening is the last 30 days.

Period controls of the References tab — **Fig. 12** · The period of this tab: dates or shortcuts.

Timeline of references per day — **Fig. 13** · One row per day, the most recent at the top, in Detail view. The hour axis is common to all days.

1: Summary of the period: references, lots, days and parts.
2: Legend. Each reference has a colour. They are never green, red or amber, which are kept for good, bad and warning.
3: The day, with the parts made and how many references went through.
4: A block is a run: consecutive snapshots of the same lot while parts were being made. A snapshot with no new parts closes the block, because it means the machine was stopped. Hover over it to see the reference, the start of the lot, the times, the duration, the snapshots, the parts and the NOK percentage.
5: View. Detail draws the runs as explained in point 4; Grouped joins them into a single block per reference. The choice is remembered in this browser.
6: Hour scale, repeated above and below.

With Grouped, consecutive runs of the same reference are drawn as a single block, even if the lot changes. Snapshots with no new parts are not drawn, and a block is only split when the reference changes or when more than 30 minutes pass without production, so real stops remain visible. Only the drawing changes; the summary, the figures for each day and the totals table are the same in both views.

Times The files carry the time on a 12-hour clock with no AM/PM. The program completes it with the file's own date, so afternoon and night appear where they belong. If the folder was copied without keeping the file dates, the time is shown as written.

Clicking a block It loads that reference and that lot into History and jumps there. It is the quickest way to find a specific lot. In Grouped view, if the block holds more than one lot, the reference is opened for that day with all its lots; hovering then shows how many lots there are.

Totals by reference in the range

Table of totals by reference — **Fig. 14** · The reference with the most parts at the top.

For each reference: days, lots, runs, parts, production time, NOK parts and NOK percentage. Clicking a row opens that reference in History, with every lot in the period.

06“Report” tab

The full report of one snapshot: everything the vision system knew about the lot at that moment.

Controls of the Report tab — **Fig. 15** · The two controls of the report.

1: Snapshot The snapshots of the History selection, the most recent at the top. If you choose none, the report always follows the last one; if you choose an older one, it stays on it.
2: Print or save as PDF Opens the browser's print dialog.

6.1 · Heading and summary

Heading and summary of the lot report — **Fig. 16** · Which snapshot it is, of which lot and which file, and the counters.

At the top, the reference, the date of the snapshot, the start of the lot, the file name and when the report was generated. Under Summary, the machine total and, separately, OK + NOK + NV with the breakdown and the percentages. If the two totals do not match, the note below gives the difference.

6.2 · Camera by camera

The rest of the report is repeated for each camera, with whichever sections that camera has.

Dimensional characteristics table of one camera — **Fig. 17** · Dimensional characteristics of one camera.

Column	What it is
Nominal, tolerance	the drawing value and the minimum and maximum allowed
n	the parts over which the vision system computed the statistics
Mean, σ	as given by the file
Cp, Cpk	as given by the file
Measured min / max	the smallest and the largest part
Below min, above max	parts below and above the tolerance
NOK counter	parts this control has rejected since the start of the lot
Status	capable if Cpk ≥ target · borderline if Cpk ≥ 1 · not capable if Cpk < 1 · beyond limit if it is negative

Histograms of the characteristics of one camera — **Fig. 18** · Histograms: one card per characteristic, with the top edge in the colour of the status.

Detail of a histogram card — **Fig. 19** · One card up close.

At the top, the characteristic, the status, and Cp, Cpk and mean. In the drawing, the bars are the histogram carried by the file; the red lines, the limits (LSL, USL); the dotted one, the nominal (N); and the amber one, the mean (μ). The pink background is out of tolerance. Below, the parts in the histogram and how many could not be measured. Hover over a bar to see how many parts it holds.

Attribute characteristics table of one camera — **Fig. 20** · Attribute characteristics: the same columns as in History.

NOK counters per control of one camera — **Fig. 21** · NOK counters per control: which control of this camera rejected the most parts in the lot.

6.3 · Lot evolution

Lot evolution, folded — **Fig. 22** · At the end of the report, folded. In brackets, how many snapshots the lot has.

Lot evolution, unfolded — **Fig. 23** · Clicking the title unfolds it: one row per snapshot.

For each snapshot of the lot: the cumulative counters and, on the right, what was made in the interval — parts, NOK and NOK percentage. The first row is what the lot already had at the first snapshot. If it is unfolded before printing, the table is included in the PDF.

07“Export” tab

1: What will be exported. Always the History selection: period, reference and lot. To change it, go to that tab.
2: Snapshots One row per file: date, reference, lot, cumulative counters, whether there was a reset and the interval increments.
3: Characteristics One row per snapshot and characteristic: limits, n, mean, σ, Cp, Cpk, measured minimum and maximum and parts out.
4: NOK per control One row per snapshot and control: cumulative NOK counter and interval increment.
5: PDF report Create PDF jumps to the report of the chosen snapshot and opens the print dialog. Choose “Save as PDF”.

The CSV files use a semicolon as separator. Decimals follow the interface language: a point in English, and a comma in Catalan, Spanish, Basque, French and Portuguese. The file name includes the folder, the period and the reference.

08The footer

Vaktari Statistic footer — **Fig. 25** · The footer is the same on every tab.

Where the data comes from: the folder, how many files were read and when the last snapshot is from.
The version, on the last line, in bold. It is the first thing you will be asked for if you have to report an issue.
If server v… appears next to it, the program and the page are not the same version. Reload the page with Ctrl + F5; if it is still there, tell whoever did the installation.

09How to read the numbers

Cumulative and interval

The counters in the file can only add up. To know what happened at a given moment you have to look at the interval. A lot with 3 % cumulative NOK may have had an hour at 15 % and the rest close to zero: the lots table shows the 3 %; the NOK and NV per interval chart, the bad hour.

Why the total does not match

The machine total is an independent counter and does not always equal OK + NOK + NV. Vaktari does not pick which one is right: it shows both, gives the difference, and uses OK + NOK + NV for the percentages so that they add up to 100 %.

Lots and resets

A lot lasts until the machine's counters are reset to zero. If the counters are reset in the middle of a lot, Vaktari will see two lots; if they are not reset when the lot changes, it will see only one. To get clean lots, the counters have to be reset when they should be, and only then.

Cp, Cpk and n

Cp says whether the parts would fit within the tolerance with the process well centred. Cpk, whether they fit as it is now. A high Cp with a low Cpk means narrow but off-centre, and is fixed with an offset; both low means too much spread.

Cpk	Status	Parts out, with the process centred
≥ 1.33	capable	63 ppm or fewer
≥ 1.00	borderline	up to 0.27 %
< 1.00	not capable	more than 0.27 %
< 0	beyond limit	more than 50 %: the mean is outside the tolerance

The 1.33 target can be changed at installation. The n of each characteristic is the number of parts the vision system used for the calculation, and it can be much smaller than the lot total: look at it before trusting a Cpk.

Three more details

Not measured are parts on which the vision system could not take the dimension. They are neither good nor bad for this characteristic.
The bell curve on the histograms is drawn with the σ derived from Cp, because the file gives σ with only two decimals. The bars are the ones from the file.
The values in the tables are from the last snapshot in the selection. To see what they were before, choose another snapshot in Report or look at the trend of the characteristic.

10Common problems

What happens	Why	What to do
The page does not open	the program is not running	Start `vaktari_statistic.bat` again and leave the black window open.
It opens on the machine's PC but not from another	the address is wrong, or the firewall is blocking port 8763	Check the address. If it is right, tell whoever did the installation.
The server cannot be reached	the program has stopped or the network is down	What is on screen is the last data received. Start the program again; the page reconnects by itself.
The folder holds no .DATA files	the path is wrong, or the folder cannot be accessed	Tell whoever did the installation: the configured folder has to be checked.
No snapshots in this selection	the period, the reference or the lot hold no data	Press All, or go to References and click a block.
Choose a reference to see its characteristics	Reference is set to All	Choose one. Characteristics of different references cannot be put in a single table.
Only one snapshot	the lot has just started	The trend needs at least two snapshots. Wait for the next one.
Many very short lots appear	the counters are reset often	Each reset starts a lot. See Lots and resets.
No new snapshots arrive	the vision system is not saving, or is saving to another folder	Look at the Last snapshot time in the footer. If it is old, check that the vision system is saving.
Excel puts the whole CSV in one column	Excel and the file do not use the same separator	Import it from Data → From Text/CSV and choose the semicolon.
The decimals in the CSV come out wrong	it was exported with the interface in one language and Excel expects another	Set the interface to Excel's language and download again.
It takes a long time to open the first time	the folder holds many files and they all have to be read	Wait for it to finish. After that it only reads the new ones.

If you have to report an issue, note the version shown in the page footer and take a screenshot of the whole screen.