A live board is a screen that shows the plant's actual current state, computed from machine signals and digitized floor data, in the place where a dry-erase production board used to hang. The whiteboard shows what someone wrote at 6 a.m. The live board shows what is true right now.

Walk most plants and you will find the same artifact at the head of every line: a gridded whiteboard with hourly targets, actuals, downtime reasons, and a marker hanging on a string. It is the most honest technology in the building, and also the most limited. This guide covers what the whiteboard gets right, where it structurally fails, and how to replace it with boards that update themselves, without losing the visibility culture that made whiteboards valuable in the first place. It is part of our paperless manufacturing series and builds on the principles in visual management.

What is wrong with the production whiteboard?

Nothing is wrong with what the whiteboard tries to do. Everything is wrong with how it has to do it. The whiteboard is a manual snapshot in a building that runs continuously, and that single fact produces every familiar failure:

It is stale by design. Someone updates it hourly, at best, between other duties. By 9:40 the 9:00 numbers are history and the 10:00 numbers do not exist yet. Decisions made off the board are decisions made off the past. It is wrong under pressure. The busier the line, the less time anyone has to update the board, so accuracy collapses exactly when visibility matters most. A bad shift produces an empty board. It has no memory. Erasing is the update mechanism. Last Tuesday's pattern, the same jam, same hour, same product, is gone, which is why the whiteboard never taught anyone anything about trends; that job fell to whoever retyped its numbers into a spreadsheet, the habit our guide to replacing Excel on the plant floor takes apart. It is one place. The board is visible from the line and nowhere else. The plant manager's view of it is a walk; the second shift's view of first shift is whatever survived the eraser.

And the quiet cost: the person updating it. Thirty minutes a shift of a lead's time spent transcribing counts that a machine already knows, times every line, every day. Our ROI calculators will put a yearly number on that transcription habit, and it is rarely small.

Information freshness over one shift: whiteboard versus live boardHow current is the board, hour by hour?freshstale6am9am12pm3pm6pmwhiteboard: decays between updates,flatlines during the crisislive board: seconds old, all shiftThe sawtooth is the manual update cycle. The gap in the middle is a bad hour, when nobody had time to write.
Manual boards decay between updates and go dark in exactly the hours that matter. A board fed by the floor itself has no sawtooth.

What is a live board?

A live board is a screen, at the line, in the hallway, on a phone, that renders the plant's real-time layer for a specific audience. The counts come from machine signals, the same source discipline as machine monitoring. The quality status comes from checks completed on tablets. The downtime reasons come from events coded when they happened. Nobody updates the board, because the board is not a document; it is a view. That distinction carries three consequences the whiteboard cannot match: it is current everywhere at once, it remembers everything (last Tuesday's pattern is a query, not a memory), and it can compute, showing pace against takt, projected end-of-shift, and OEE from source signals rather than estimates, the calculation our OEE guide walks through.

A live board is also where escalation starts rather than where it ends. On a whiteboard, a red hour is a note about the past. On a live board wired into an andon system, a red state is an event: it can page the right person, start the clock on response time, and log the resolution. The board stops being a report and becomes an interface.

What belongs on a live board?

Less than you think, chosen per audience. The whiteboard's genius was enforced brevity, and screens tempt people into dashboard clutter. The working pattern is three tiers on one data layer:

Line board: for operators, at the line. Current run, pace against target, quality status, active issue and who owns it. Big type, readable at ten meters, glanceable in three seconds, the same rules good visual management always enforced. Supervisor board: all lines on one screen, exceptions first: what is behind, what is down, what is waiting on a decision. This replaces the walking tour of whiteboards and most of the morning spreadsheet, because it is production reporting rendered continuously instead of compiled nightly. Plant board: for leadership, the day against plan, the week's trend, and the three problems worth attention. If a tier's board needs a legend, it is carrying numbers that belong on a different tier, or in the drill-down, not on the glass.

Three board tiers on one real-time layerOne layer, three audiencesmachine signalscounts, states, ratesdigitized floor datachecks, downtime codes, holdsREAL-TIME LAYER: one current model of the plant, full history retainedLINE BOARDrun, pace, quality,active issue + ownerSUPERVISOR BOARDall lines, exceptionsfirst, one screenPLANT BOARDday vs plan, trend,three problems that matterSame numbers at every tier, because every tier reads the same layer. No reconciling.
Three audiences, three boards, one source. The supervisor's number and the operator's number can no longer disagree.

How do you replace whiteboards with live boards?

The order of operations decides whether the screens get trusted or ignored. Plants that bolt a TV to the wall and call it digital get a very expensive whiteboard. The sequence that works:

  1. Photograph every whiteboard in the plant. They are a map of what the floor already decided it needs to see. The live boards' first job is to show exactly this, not what a dashboard template ships with.
  2. Digitize the inputs first. A live board is only as live as its sources. Move the hourly counts, checks, and downtime codes to tablets at the station, and connect machine counts where PLCs make it cheap. Screens before sources is the classic failure.
  3. Stand up the line boards with the crews. Build each line's board next to the whiteboard it replaces, with the operators who used it. Run both for two weeks; retire the marker when the crew, not the project team, says the screen wins.
  4. Add the supervisor and plant tiers. Exceptions first, trends second, vanity metrics never. These tiers are views on the same layer, so they cost days, not projects.
  5. Wire in escalation. Connect red states to notification so the board acts like an andon, not a poster. Response time becomes a measured number.
  6. Keep one whiteboard. Seriously. Problem-solving, kaizen sketches, and the day's oddities still belong on a surface anyone can scribble on. The goal was never to eliminate markers; it was to stop using humans as data buses.

Do you lose anything when the whiteboard goes?

Honestly: you can, if you digitize carelessly. The whiteboard had virtues worth defending. It was owned by the crew, not by IT. It was writable by anyone in three seconds. And updating it, at its best, was a small hourly act of accountability. A live board rollout should preserve all three: crews configure their own line boards, every board keeps a fast human-input path (notes, flags, shout-outs), and the hourly rhythm survives as a brief crew huddle at the screen rather than a transcription chore. What you should refuse to preserve is the transcription itself, and the fiction that a hand-copied number at 6 a.m. described the plant at 2 p.m. The handover between shifts deserves the same upgrade, which is its own topic: see digitizing shift handover.

The measurement standards back the shift. A few primary anchors:

How do live boards fit the bigger picture?

Boards are the visible tip of the layer underneath. Once machine signals, digitized forms, and schedules feed one real-time model, the same layer that renders the boards can also act on what they show, drafting the work order for the recurring 10 a.m. jam, flagging the material shortage before the line starves, compiling the shift report nobody has to type. That acting layer is the subject of our flagship guide, what is an AI-native MES. This is how Harmony AI deploys it: digitize the paper, connect the machines, then light up role-specific boards and apps on the same data model, in weeks, with no rip-and-replace. The CLS case study shows the result on a real floor: dashboards and automated reporting running against a single source of truth.

The bottom line

The whiteboard earned its place by making the day visible. Replace it with something that keeps the visibility and drops the manual labor: boards rendered from live floor data, tiered by audience, wired to escalate, with history that survives the eraser. Keep one whiteboard for thinking. Retire the rest, and give the lead back the half hour a shift they spent being a photocopier.