Which fields should a per-shift quality report include?

Which fields should the per-shift report have?

The report has to answer three questions: how many eggs passed, how they were classified and why. The minimum fields are these:

• Date, shift, line and responsible team.
• Total eggs for the shift and the calculation base of the rates.
• The count of each class: clean, dirty, cracked and unverified.
• The rate of each class, with the unverified rate highlighted.
• The three most frequent reasons within each defect class.
• Notes from the team, such as stoppages, a batch change or a camera adjustment.

How do you count the classes without counting an egg twice?

Each egg has one class only in the Vigiovo record. So the sum of clean, dirty, cracked and unverified eggs must match the shift total. If it does not, there is an egg without a record or a duplicated record, and the report should not be closed before the cause is found.

Before comparing shifts, confirm how the configuration handles an egg that is both dirty and cracked. The egg takes one class, and the readable reason in its record shows the measures that went over the limit. Use the same rule every time, or the comparison between shifts means nothing.

What do the reasons and the unverified eggs tell you?

The reasons in each record are the most useful part of the report, because they say which measure went over its limit: the dirty fraction of the shell, or the crack length as a multiple of the egg radius. A shift with many dirty eggs from the dirty fraction has a different problem from a shift with long cracks.

An unverified egg gets no grade and goes to manual review. The report should show that rate separately, with its most common reasons. The table gives the practical reading of each reason.

How often should the report be reviewed?

Review the report at the end of every shift, while the team still remembers what happened on the line. The review is short and follows this order:

1. Check that the sum of the classes matches the shift total.
2. Calculate the unverified rate and compare it with the previous shift.
3. Read the three most frequent reasons in each defect class.
4. Record the action taken, if any, and the time of the review.

Once a week, compare the shifts with each other and with the previous week. One shift on its own may be a one-off. A pattern that repeats across different shifts usually points to the process, not to the people who were on the line.

What does the per-shift report not show?

The report measures the shell the camera can see. It does not show the air cell, the yolk, internal spots or hairline cracks, which need candling or acoustic testing.

So the report should be read as the result of the camera stage, not as the final result for the batch. When a reason keeps repeating, use the per-egg records to investigate, rather than relying on the summary alone.

Frequently asked questions

Does a per-shift report have to be built in a spreadsheet?

No. The report can come from a dashboard or from an export of the per-egg records, as long as it has the minimum fields and the same calculation base in every shift. The format matters less than the consistency of the numbers.

Which indicator should I look at first when I open the report?

First, check that the class counts add up to the shift total, because that confirms the report is complete. Then look at the unverified rate, which shows whether the camera can see the line. Only after that read the dirty and cracked rates and the most frequent reasons.

Who should review the report?

The person responsible for the line, together with the team that worked the shift. The review should be short and done right after the shift ends, so the team's observations are still fresh and the record reflects what really happened.

Sources

1. Como o Vigiovo classifica ovos sem IA / How Vigiovo grades eggs without AI, Vigiovo