Countdown

The usual workaround is a measure such as DATEDIFF ( MAX ( Incidents[Date] ), TODAY (), DAY ) on a Card. It has two faults. A report page does not re-run its queries on its own, so a page left open on a wall screen shows the number from when it was loaded until someone interacts or the model refreshes. And in the Power BI Service, TODAY () and NOW () run in UTC, so a plant in Central Europe sees the day change at 01:00 or 02:00, and a plant in North America sees it in the early evening.

SmartVisuals Countdown & Days-Since computes the count in the browser, from the viewer's clock, and keeps ticking between dataset refreshes. It ticks once a minute when seconds are not on screen and once a second when they are, stops entirely when the visual is hidden or scrolled out of view, and recomputes from a monotonic clock after a laptop sleeps, so it never drifts. Where a device clock cannot be trusted, bind a measure such as UTCNOW () to Reference time and a device clock running behind the model is corrected.

A Power BI datetime carries no time zone: 2026-12-31 23:59 is just a wall-clock reading. Left alone, it resolves against the zone of whoever opens the report, so the same deadline arrives at different moments in Prague and in Chicago. The Time Zone card makes the choice explicit: the report viewer's local zone, UTC, a fixed offset such as +02:00, or a named zone such as Europe/Prague. With a named zone, a shop floor ends its day at its own midnight on every screen.

Daylight saving follows a stated policy. A time that does not exist on a spring-forward night is moved forward by the length of the gap, and a time that occurs twice on a fall-back night resolves to the earlier occurrence. Both are tested against Python's zoneinfo in 210 cases, including half-hour and quarter-hour zones and southern-hemisphere transitions. Show The Zone On The Visual prints the zone and its current offset under the number.

Sixty calendar days to an audit is the wrong number when the plant shuts for a week in between. Turn on Count Working Time, pick the working weekdays, and bind a column of public holidays and shutdown dates to Non-working dates. The countdown keeps its calendar reading and adds a line with the working days left. Limit To Business Hours narrows the count to a daily window, including one such as 22:00 to 06:00 for a night shift.

The working-day counts are tested against numpy.busday_count with real holiday sets, 84 cases in all. The holiday table does not need a relationship to the deadline table; binding both gives the visual every distinct holiday date and each milestone still appears once.