BRENNEN

Industries / Construction and property / Commercial and residential construction / Painting

Color

This is the work of choosing, naming, specifying and matching the colors a building is painted, from a color card or a scanned sample to the tinted can.

4

Milestones achieved, from 1979 to 2024.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1979

    A standard for seen color

    The Natural Colour System (NCS) describes a color by how people see it, by its hue, its blackness and its chromaticness (how strong the color is), building on Ewald Hering's opponent-process theory of vision, rather than by the pigments that make it. Researchers at the Swedish Colour Centre Foundation worked it up from the 1960s, it was launched in 1979 with an atlas of standard colors, and in 1981 Anders Hård and Lars Sivik described it as a Swedish standard for color notation. NCS Colour's own history gives both 1,412 and 1,450 for the first atlas and 2,050 colors today; South Africa uses it as a national standard, and ASTM E2970 covers it in the United States.

    ncscolour.comdoi.org

    Credit

    Anders Hård, Lars Sivik and Gunnar Tonnquist

  2. 2014

    Walls recolored on a phone

    AkzoNobel, the maker of Dulux, launched its Visualizer app in 2014. It lets a user see their own walls in different colors in real time, before buying paint. By June 2018, AkzoNobel counted more than 20 million downloads in more than 60 markets and over 400 million visualizations, its own figures.

    akzonobel.com

    Credit

    AkzoNobel

  3. 2015

    Contrast measured as reflectance

    Light reflectance value (LRV) is the share of visible light a surface reflects. In England's building guidance for access to buildings other than homes, Approved Document M volume 2, in force from 1 October 2015, two surfaces contrast visually when their LRVs differ by more than 30 points, or by at least 20 points where more than 200 lux of light falls on them. It asks for such contrast on items such as internal doors, door furniture, seating and sanitary fittings, so color choices for these buildings have a number to meet.

    assets.publishing.service.gov.uk

    Credit

    UK government, citing design guidance from the University of Reading

  4. 2024

    Pocket color readers scored

    To match an existing wall, a painter or shop measures its color and looks up the nearest paint. Researchers at the University of Ljubljana, the University of Latvia and veterinary science firms in Slovenia and Latvia tested four handheld color readers priced at US$100 to 1,200, and a smartphone camera, on 183 colors of the RAL Design System Plus paint chart. The Nix Spectro 2 matched 99% of the colors, the Spectro 1 Pro and Datacolor ColorReader about 85%, and the Pico reader and the phone 54 to 77%.

    Lessonsuch results age fast, since a 2021 printing-industry study had found the previous low-cost models (Nix Mini, Nix Pro 2, Variable Spectro 1) less precise and consistent than professional instruments.

    pmc.ncbi.nlm.nih.govprinting.org

    Credit

    Jaša Samec, Blaž Cugmas and colleagues

Watch for

  1. An AI agent picks colors

    Behr Paint Company says it combined its color data with Google Cloud's Gemini models and Vertex AI to build ChatHUE, an AI agent that guides buyers to a paint color. Behr cites a study it commissioned: more than 40% of consumers find choosing a color the hardest step in a paint project, and 40% say indecision stops them starting.

    Achieved whenan independent study finds that buyers who use it choose a color sooner, or repaint less, than buyers who do not.

    googlecloudpresscorner.com

    Claimed by

    Behr Paint Company and Google Cloud, July 2025

Open

  1. 01

    Check AI color advice

    A public test of whether AI color assistants give advice that meets a measurable rule, starting with England's 30-point LRV contrast between a door and its wall, run on 100 written briefs.

    The brief
    Credit · unclaimed
  2. 02

    Check visualizer colors against paint

    A measured test of how closely the colors that paint visualizer apps show on a phone screen match the same colors painted on a wall in the same room.

    The brief
    Credit · unclaimed

Open 01 · the brief

Check AI color advice

Credit · unclaimed

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

What to build

A public test of whether AI color assistants give advice that meets a measurable rule, starting with England's 30-point LRV contrast between a door and its wall, run on 100 written briefs.

Why it is still open

Behr describes ChatHUE as guiding buyers to a color they like, and Sherwin-Williams says its Color Expert app suggests colors from those already in a room, such as art and furniture; neither mentions contrast rules, and a Crossref and GitHub search found no published test of such advice against one.

pcimag.comassets.publishing.service.gov.uk

Built with what exists

Behr's ChatHUE, Sherwin-Williams' Color Expert and a general model such as Claude, each given the same briefs; the LRV the paint maker lists for each suggested color, or a reading from a handheld spectrophotometer where none is listed; a spreadsheet to score each answer (see the Note "A thousand questions, one pass").

  1. First step

    Write 20 briefs for a care home corridor and an office, each naming a wall color, and ask each assistant for a door color.

  2. Done looks like

    100 briefs, each put to at least three assistants, with the share of answers that meet the 30-point rule, and every prompt and answer published.

Open 02 · the brief

Check visualizer colors against paint

Credit · unclaimed

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

What to build

A measured test of how closely the colors that paint visualizer apps show on a phone screen match the same colors painted on a wall in the same room.

Why it is still open

AkzoNobel reports 20 million downloads and 400 million visualizations for its app but no figure for how true its colors are, and a Crossref search found no study of the color accuracy of paint visualizer apps. The independent tests found measure handheld readers and phone cameras on paint charts, not what the apps display.

akzonobel.compmc.ncbi.nlm.nih.gov

Built with what exists

Dulux Visualizer and Sherwin-Williams' ColorSnap Visualizer; sample boards painted in 30 colors; a handheld spectrophotometer such as the Nix Spectro 2; a display colorimeter to measure the phone screen; colour-science, an open-source Python library, to compute ΔE00, a standard number for how far apart two colors look.

  1. First step

    Paint five sample boards, measure them, and measure each app's rendering of the same five colors on one phone in one room.

  2. Done looks like

    30 colors shown in two rooms under daylight and warm LED light, with the ΔE00 between each app's screen and the painted board for every color, published with photos and the method.