Digital Production Boards, TV Dashboards & Hourly Alert Systems

The traditional sewing floor learns about yesterday in a morning meeting. Supervisors report output from memory and paper sheets, the meeting runs long, and lines that struggled in the morning receive help only after the meeting ends. The information is honest but late, and late information cannot recover a lost morning. A digital production board changes the timing: the floor learns about the current hour while the next hour is still available to fix it.

A TV dashboard on each line shows the same six figures all day: hourly target, hourly actual, cumulative variance, line efficiency, first-pass quality, and absenteeism. The production manager sees all lines at a glance instead of collecting verbal reports. The supervisor sees a struggling line the hour it struggles. The operators see the target they own, in numbers large enough to read from the end of the line. The board does not manage the floor. It gives the floor the information to manage itself, hour by hour.

Six figures earn a place on the board, and the discipline is in refusing the seventh. Hourly target states what the line owes this hour. Hourly actual states what the line produced. Cumulative variance states where the day stands: total actual minus total target from the start of the day. Line efficiency states how much of the attended time became standard earned minutes. First-pass quality states what share of garments passed inspection without repair, because output without quality is rework, not production. Absenteeism states how many planned operators are missing, because a line short of hands cannot meet a target set for a full line.

Every other figure is analysis, not board material. Style-wise output, operator-wise output, defect Pareto charts, and weekly trends belong in the IE office, reviewed daily or weekly, not on a screen the line watches all day. A board crowded with twenty figures guarantees that nobody can tell at a glance whether the line is winning or losing, and a board that cannot be read at a glance will not be read at all. Six figures, large type, facing the line, updated every hour: that is the whole design.

The rhythm is simple and it must not slip. Within five minutes of each hour ending, the hourly actual is entered and the board recomputes variance and cumulative position. The line leader reads the variance and names the cause for that hour: absenteeism, machine breakdown, material starvation, or quality rework. If the hour is green, at 90 percent of target or above, the line continues. If the hour is yellow, below 85 percent, the line leader investigates at the next natural break. If the hour is red, below 75 percent, the supervisor visits the line within fifteen minutes with a named action.

Consider one line with an hourly target of 80 pieces across eight hours, a daily target of 640 pieces. Hour 1 produces 62 pieces, a variance of minus 18, or 77.5 percent of target: a red hour, and the supervisor visits within fifteen minutes. Hours 2 to 4 produce 78, 85, and 80 pieces, bringing cumulative actual to 305 against a cumulative target of 320. Hours 5 to 8 produce 74, 88, 90, and 82 pieces, bringing the day to 639 pieces against 640, or 99.8 percent. The day was recovered hour by hour after a red first hour, and that recovery was possible only because the shortfall was visible in hour 1, not discovered at the end of the day.

Alert thresholds must be agreed before going live, and written down with the action each level triggers. Green, at 90 percent or above, needs no action. Yellow, below 85 percent, asks the line leader to investigate at the next break. Red, below 75 percent, sends the supervisor to the line within fifteen minutes. Two consecutive red hours escalate to a joint action by IE and the production manager, immediately. The thresholds matter less than the discipline: an alert without a named owner and a response time is decoration, visible to everyone and owned by nobody.

Alert fatigue is the failure mode to design against. If thresholds are so tight that every hour is red, the team learns within a week to ignore the board, and the board becomes wall decoration. Red must be rare enough to mean something. When a board shows red all day, the first suspect is the target, not the line: a target set without honest SMV and efficiency assumptions will punish good lines all day. Review the target before scolding the line, and keep red as a genuine call for help.

The daily huddle replaces the hour-long morning meeting. Ten minutes, standing at the dashboard, three questions only: what went red yesterday, what fixed it, and what is different today. Lines that stayed green are not discussed. Individual performance is never discussed in the huddle; it is handled privately. The huddle ends with named actions for today red risks: a machine to be serviced before start, a material delivery to be chased, an absent operator to be covered.

A factory in Dhaka that made this change cut its morning meeting from about 45 minutes to about 10 minutes in the illustrative pilot period, and red lines received help within fifteen minutes instead of after the meeting. The time saved was not the main gain. The main gain was that help arrived while the day could still be recovered, instead of arriving as a report about a day already lost.

Choose six figures and refuse the seventh. Mount the screen facing the line at eye level, with numbers readable from the end of the line. Agree thresholds and escalation rules in writing before the first live day. Enter actuals within five minutes of each hour ending, every hour, without exception. Escalate red within fifteen minutes with a named action. Huddle for ten minutes daily and stop meeting about yesterday. Review thresholds monthly: if red fires constantly, fix the target or the threshold; if red never fires, the threshold is too loose to protect the shipment. A board run this way becomes the floor management system within weeks, not because the screen is impressive, but because help arrives while it can still change the day.

A production board needs an owner, and the owner should be the production manager, not the IT team or the IE team alone. The owner checks three things every week: that hourly actuals were entered within five minutes of each hour ending, that red alerts received their fifteen-minute response, and that the ten-minute huddle actually happened. A board whose actuals arrive late is a board whose variances describe yesterday, and a huddle that slips back into an hour-long meeting is a board whose discipline has already decayed.

Once a month, the owner reviews thresholds and targets together. If red fired on more than about one hour in five across the floor, either targets are unachievable or thresholds are too tight, and the review must say which. If red never fired at all, thresholds are too loose to protect shipments. Targets themselves must be revisited whenever SMV values change, when a new style with a different operation balance starts, or when planned efficiency assumptions change. A board reviewed monthly stays honest. A board installed and forgotten becomes wall decoration within a quarter, and the floor will quietly return to verbal reporting while the screens keep glowing.