BRENNEN

Industries / Construction and property / Janitorial and facility services

Quality inspection

Quality inspection checks that cleaning was done and done well, by looking, by measuring dirt and germs on surfaces, and by auditing against a standard that customer and contractor agree on.

7

Milestones achieved, from 1994 to 2021.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1994

    A meter for surface dust

    Thomas Schneider and colleagues in Denmark published an instrument method for measuring the dust left on a surface, so that a customer could check whether it was getting the cleaning it paid for, as Schneider and his co-authors explained in 2002. It later became the optional instrument method of the Nordic cleaning quality standard INSTA 800.

    irbnet.dedoi.org

    Credit

    Thomas Schneider and colleagues, National Institute of Occupational Health, Copenhagen

  2. 2000

    Looking clean is not clean

    Christopher Griffith and colleagues tested up to 113 surfaces in a British operating theatre and hospital ward three ways: by eye, by growing bacteria from swabs, and by ATP bioluminescence, a swab test that measures adenosine triphosphate, a molecule found in living cells and food residue, by the light it gives off when mixed with an enzyme. After routine cleaning, 70% of sites failed the microbiological limits and 76% the ATP limits, but only 18% looked unacceptable.

    LessonVisual inspection alone passes most surfaces that measured tests fail.

    doi.org

    Credit

    Christopher Griffith, Rose Cooper and colleagues, published in the Journal of Hospital Infection

  3. 2000

    Counting dirt, not tasks

    INSTA 800, a common Nordic standard published through Danish Standards in 2000, judges a cleaned room by what is left on it rather than by which tasks were done. Inspectors count four kinds of soil (litter and loose dirt, dust, stains, and surface dirt) on furniture, walls, floors and ceilings, compare the counts with five quality levels, and inspect a random sample of rooms using the ISO 2859-1 sampling tables. Before, contracts listed tasks and how often to do them. It follows EN 13549 (2001), the European standard for cleaning quality measuring systems, became certifiable in 2010, and was split into two parts in 2018; SINTEF, a Norwegian body, certifies its inspectors.

    irbnet.deds.dksintefcertification.no

    Credit

    INSTA, the Nordic standards bodies' joint program; Danish Standards

  4. 2004

    Cleaning tested like a drug

    In Oslo in 1998, researchers chose 114 office workers who had reported irritated eyes, nose or throat, gave some of their offices a thorough cleaning and the rest a light cleaning as a placebo, and did not tell the workers or the people measuring them which was which. Airborne dust in the cleaned offices fell from 67 to 50 µg/m³. Workers in those offices were 3.5 times as likely to report a large drop in irritation and had less nasal congestion, measured with an instrument.

    doi.org

    Credit

    Kari Skulberg, Kristian Skyberg and colleagues, published in Epidemiology in 2004

  5. 2008

    Fluorescent dots in hospitals

    Philip Carling's group put small dots of a gel that is invisible in normal light but glows under ultraviolet light on high-touch surfaces in patient rooms before cleaning, then checked with a UV lamp which dots had been wiped away. Across 36 US hospitals, only 48% of 20,646 marked surfaces were cleaned at first. After training and repeated feedback to cleaning staff, 77% of 9,464 surfaces were.

    doi.org

    Credit

    Philip Carling and the Healthcare Environmental Hygiene Study Group

  6. 2017

    Cleaner rooms, same infections

    A randomized trial in 16 hospitals in northeastern Ohio (15 finished) added monitoring with fluorescent markers and feedback to cleaning in half of them. More markers were removed and less Clostridioides difficile, a gut bacterium that spreads through surfaces in hospitals, was found on surfaces, but infections acquired in hospital did not fall.

    LessonA better cleaning score does not by itself show fewer infections, so programs that aim to cut infections have to measure infections too.

    doi.org

    Credit

    Amy Ray, Curtis Donskey and colleagues, published in Infection Control and Hospital Epidemiology

  7. 2021

    Star ratings on the wall

    NHS England's National Standards of Healthcare Cleanliness 2021 require hospitals to put every area into one of six functional risk categories, which set how often it is cleaned and audited. Efficacy checks, in which an auditor watches cleaning being done, are required at least yearly for the four highest-risk categories, with a minimum score of 80%. Cleaning scores must be shown where the public can see them: percentages for six months, then star ratings from zero to five, over an 18-month changeover.

    england.nhs.uk

    Credit

    NHS England and NHS Improvement

Watch for

  1. A phone grades the dots

    A four-campus hospital system in Southwest China reports that a smartphone app, which compares ultraviolet photos of fluorescent marker dots taken before and after cleaning, raised its weekly marker removal rate from 83.91% in the first week to above 90% from about week four, over 2,457 markers in 17 weeks from June to October 2025. The spread between weeks also narrowed, from a standard deviation of 29.72 to 9.59 percentage points. The study is the hospital's own, and its abstract reports no check of the app's readings against trained human readers.

    Achieved whenthe app's calls are compared with trained human readers on the same markers, and a hospital outside the authors' system reports similar results.

    doi.org

    Claimed by

    Wang P. and colleagues, Antimicrobial Resistance and Infection Control, August 2026

Open

  1. 01

    Agreement between inspectors

    A published test of how often certified INSTA 800 inspectors reach the same verdict on the same rooms. Several inspectors score each room independently right after cleaning, and their counts of litter, dust, stains and surface dirt, and their pass or fail calls, are compared, with instrument dust readings as a check.

    The brief
    Credit · unclaimed
  2. 02

    How fast rooms get dirty

    Measured re-soiling curves: how quickly dust, litter and stains return to offices, classrooms and corridors after cleaning, set against the number of people who used each room, so cleaning frequency can be set from data rather than habit.

    The brief
    Credit · unclaimed

Open 01 · the brief

Agreement between inspectors

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A published test of how often certified INSTA 800 inspectors reach the same verdict on the same rooms. Several inspectors score each room independently right after cleaning, and their counts of litter, dust, stains and surface dirt, and their pass or fail calls, are compared, with instrument dust readings as a check.

Why it is still open

INSTA 800 rests mainly on visual counts, with instruments optional, and SINTEF's certification tests that inspectors know the standard, not that two of them score a room alike. Searches of the web, Europe PMC, Crossref and GitHub ("INSTA 800" with "agreement", "inter-rater", "reliability") found no published agreement study. Griffith's 2000 hospital study shows how far a visual judgment can drift from a measured one.

ds.dksintefcertification.nodoi.org

Built with what exists

The INSTA 800-1:2018 standard, sold by Danish Standards; three to five certified inspectors; rooms of several types; a surface-dust measuring method of the kind the standard allows as an option; a phone photo of each counted defect; Fleiss' kappa, a standard score of agreement among several raters, computed with Python's statsmodels library.

  1. First step

    Ask one cleaning company that works to INSTA 800 to have two of its certified inspectors score the same ten rooms on the same morning, then compare the sheets.

  2. Done looks like

    At least 100 rooms of at least five types, each scored independently by at least three certified inspectors, with kappa for each kind of soil and the share of rooms where pass or fail calls differ, published with the photos and dust readings.

Open 02 · the brief

How fast rooms get dirty

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

Measured re-soiling curves: how quickly dust, litter and stains return to offices, classrooms and corridors after cleaning, set against the number of people who used each room, so cleaning frequency can be set from data rather than habit.

Why it is still open

INSTA 800 judges a room immediately after cleaning and does not address how fast it gets dirty again, as its users noted in 2002. Studies have measured dust before and after a single cleaning, as in Oslo in 1998, and Croxford and colleagues sampled surface dust weekly at desks and walkways in a London office in 2000, to test air filters rather than cleaning, but searches of the web, Europe PMC and Crossref ("re-soiling", "recontamination", "dust accumulation" with "office" and "cleaning frequency") found no published per-day re-soiling rate tied to measured occupancy.

irbnet.dedoi.orgaivc.org

Built with what exists

Door counters or low-cost people counters; a surface-dust measuring method; time-lapse phone photos of floors and desks; an image model that counts litter and stains, checked by a person against INSTA 800's counting rules; Python for the curves.

  1. First step

    On one office floor, take dust readings and photos each morning for two weeks in five rooms, with a simple door counter on each room.

  2. Done looks like

    Daily readings in at least 30 rooms of three types over eight weeks, giving for each room the days until it drops one INSTA 800 or APPA level, plotted against people per day and published with the data.