Home
Resources
The log says everyone washes their hands. What cameras showed at food plants

The log says everyone washes their hands. What cameras showed at food plants

6 min read · Updated 2026-10-28

For: owners and managers of food plants, heads of HACCP, process engineers

Summary

At a UK pie manufacturer, 24 hours of video showed correct handwashing before entering production in only 2% of cases.
A US study of 321 food service workers: hands were washed properly in only 27% of the situations where it was required.
Workers wearing gloves washed their hands half as often as those without gloves.
While an auditor was visible nearby, hand hygiene events were about 2.5 times more frequent than when no auditor was present (3.75 vs. 1.48 per hour). This is why one-off checks show a better picture than the real one.
Continuous monitoring with regular feedback raised compliance in a hospital unit from 6.5% to 82% and kept it there for more than a year.

For: owners and managers of food plants, heads of HACCP, process engineers

What the log sees and what the camera sees

At most plants, handwashing is checked like this: there is an instruction by the sink, a log that people sign, and a head of HACCP who sometimes stands at the hygiene station. On paper, everything is fine. Researchers from Cardiff Metropolitan University looked at what it looks like when nobody is standing nearby.

Site	What was recorded	Result
Ready-made pie manufacturer, UK (2018)	24 hours of video, 674 entries into production	Correct handwashing in only 2% of cases
Food manufacturer, high-risk and high-care zones (2020)	15 hours of covert camera footage, 200 handwashing attempts	More than 98% did not meet requirements
Sandwich factory (2024)	16 hours, 12 workers, 588 situations where hands had to be washed	In 32% of cases hands were not washed at all

These are ordinary British food plants that agreed to take part in the research.

The picture in food service is similar

The US network EHS-Net, coordinated by the Centers for Disease Control and Prevention (CDC), observed 321 kitchen workers. On average there were 8.6 situations per hour when hands had to be washed. Workers attempted to wash their hands in 32% of them and washed them properly in 27%.

The CDC also notes that after touching their face or body, about one worker in ten washed their hands, and after handling raw meat or dirty equipment, about one in four. According to the CDC, in nine out of ten outbreaks where a worker contaminated the food, the route was through the hands.

For us, the most interesting part is gloves. In the same study, workers wearing gloves attempted to wash their hands in 18% of required situations, and workers without gloves in 37%. A person in gloves feels protected and skips washing, even though gloves carry the same contamination as hands.

Why an inspection sees a better picture

When people know they are being watched, they behave differently. This is called the Hawthorne effect, and it has been measured for hand hygiene. In a Canadian hospital, an electronic system counted every press of the hand sanitizer dispenser for 8 months. While a visible auditor was on the unit, the dispensers were used 3.75 times per hour. On similar units with no auditor present, the figure was 1.48 times per hour.

So a one-off check mostly measures how people behave during the check. The head of HACCP stands at the entrance and everyone washes their hands. He leaves, and the rate drops several times over. The log will not show this, because people sign it the same way in both cases.

What continuous monitoring with feedback changes

The most convincing result again comes from healthcare, but the mechanism is the same. In an intensive care unit in New York, cameras watched every sink and dispenser. For the first months the recordings were reviewed, but workers were told nothing: correct hand hygiene happened in only 6.5% of cases, 3,933 of 60,542.

Then the results started being shown to the team regularly. Compliance rose to 81.6%, and after 75 weeks it held at 87.9%. The cameras and the people stayed the same; the only change was that every case was now noticed and discussed.

Two conclusions follow for food production. First, a covert camera changes nothing by itself. It only shows what really happens. Second, what works is the combination of continuous monitoring and feedback, where every violation reaches the person responsible and is discussed with the team.

How we see this in practice

We ran video monitoring of hygiene at a meat processing plant. The cameras were already in the production areas, and the system used them to check hair caps, face masks, beard covers, and gloves. Each detection became an event card with a frame for the head of HACCP. He made the decision, the violation log built itself, and it could be shown to an auditor.

Handwashing was not on that list. We set it up for each specific site: where the sink is, which camera sees it, what counts as washing. The rule runs in shadow mode first, and we only talk about results once your own footage shows how many detections are real.

The benefit is the same as in the studies: the person responsible sees what happens during a shift when he is not on the floor himself, and talks to the team about a specific frame instead of a vague impression.

Where to start at your plant

Identify the points where handwashing is mandatory: entry into production, moving from a raw zone to a clean zone, returning after a break.
Check whether at least one camera covers these points.
Watch a few hours of ordinary footage without warning people. The number may be an unpleasant surprise, as it was for the British manufacturers.
Decide who reviews violations with the team and how often. Without this step, as the hospital experience shows, nothing will change.

We can connect to the cameras you already have and start in shadow mode, so you see the number on your own footage before anyone makes promises.