The dashboard: reading the network's state in thirty seconds
The network's state on the left, what got done over the period on the right, a per-trail table, and exactly what each number means.
The dashboard is the web app's first page. It isn't for working: it's for knowing where things stand before deciding what to do with your week.
It has two columns. On the left, the network's state: what's still on the table, trail by trail. On the right, the period: what the crew got through between two dates. One date range, top right, drives all of it.
Where: left navigation bar, under General → Dashboard. It's the page that opens when you log in.
The principle: one snapshot a day
This is the point to understand before anything else, because it explains every surprise.
The dashboard doesn't read your data live. Every night Sentinel compiles a snapshot of the day that just ended — a frozen picture — and it's that series of snapshots the dashboard displays.
Two consequences:
- Today is never shown. The range stops at yesterday at the latest. A task completed this morning shows up tomorrow.
- The numbers don't move while you work. This isn't a real-time dashboard, and that's on purpose: you compare whole days, not moments.
The page subtitle always states the last compiled day. If a night was skipped, that's where you'll see it.
To see the true state right now, open the Tasks and Structures sections. The dashboard is for seeing a trend.
The date range and the four balances
The header and the network status strip.
- 1The date range drives everything below it.
- 2The line follows the value day by day across the range; the dot marks the last day.
- 3The badge compares with the previous period of the same length, named in plain words.
By default the range covers the last seven days, ending yesterday. It's neither the calendar week nor the current month: it's a rolling window.
You can change it freely, up to a year. The future and today are blocked in the calendar, for the reason above.
Changing the range changes everything: both columns and the trail table.
The Network status strip carries four numbers:
| Indicator | What it is |
|---|---|
| Critical tasks | Tasks with Critical priority that aren't done |
| Open tasks | Every task that isn't done |
| Structures at risk | Structures at high or moderate risk |
| Overdue reports | Inspections whose due date has passed |
These are balances, never sums: each is the state on the last day of the range, and the small "as of 16 Sep" tag on the right says so. Making the range longer doesn't inflate them. Everything that adds up over the range lives in the right column.
The line and the badge
The small line behind the number traces its value day by day across the chosen range. It gives no scale: it gives a shape. A steadily rising line on "Open tasks" is a signal, even when today's number looks acceptable.
The coloured badge compares with the previous period of the same length: seven days against the seven before, a month against the month before. It shows the raw difference, not a percentage, and the comparison period is written next to it: "vs 3 Sep – 9 Sep".
Its colour follows the meaning of the news, not the sign of the number: for these four balances, going up is bad, so red. When the difference is zero, the badge gives way to "No change".
The breakdown
The breakdown on the left, the period on the right.
- 1Every bar carries its legend: the colour, the label and the count.
- 2The right column contains nothing but totals over the range.
- 3The hours trend has its own window, eight weeks, whatever the range.
Below the four balances, the Breakdown card splits what's still open into four bars, one per family:
- Open tasks by status — how many are waiting, how many are running;
- Open tasks by priority — how serious what's left is;
- Structures by risk level, including those not evaluated;
- Projects by status.
Each bar's total is written on the right, and every segment has its label and count in the legend: no hovering needed to read it. Hovering a segment still gives its card: the label, the count and its share of the total as a percentage.
Two cards complete the state: Open tasks by type, which says what kind of work dominates — useful for a funding application — and Project progress for active projects.
Where it happens
Trail health, sorted by load.
- 1Open tasks, by status.
- 2Structures, by risk level.
- 3The trail's load: its total of open tasks, which orders the table.
This is the most useful widget on the page, and it closes the left column.
Each row is a trail; on the left its open tasks split by status, in the middle its structures split by risk, on the right its load — its total of open tasks. The rating symbol recalls the trail's difficulty.
The table is sorted by load: the trail at the top is the one with the most work ahead of it. This table answers the question no other screen asks: where is the work concentrated? A trail with seven tasks waiting and two high-risk structures isn't an open trail, whatever its published condition says. It's the right starting point for setting up a project or planning a work day.
A value of zero shows as pale grey rather than a badge: your eye only sees what isn't zero.
Over the period
The right column speaks only of what happened between the two dates of the range. Here, double the duration and the numbers grow: they're counts, not balances.
- Completed tasks: the total over the range, with its comparison badge — for this one, up is green — and the day-by-day detail.
- Hours worked: the total of hours logged over the range, then hours per member. The trend chart next to it has its own window — the last eight weeks, regardless of the range you picked — and says so under the bars. Hours logged without an account (machinery, contractors) form a "Without an account" row.
- Labour cost, split by work role. This block only appears with the Costs right; its badge is grey, because a rising cost is neither good nor bad news in itself.
- Top task completers and Top task creators. The first measures execution, the second vigilance; they're almost never the same people, and that's fine.
- Reports received from the public, with those closed and refused, then receptions day by day. This block only appears if public reports are enabled for your place.
A block with no data neither disappears nor shows a message: it keeps its shape, empty — grey tracks, axes without bars, dashes where names would be — and its number turns grey. The page always keeps the same silhouette, and what's empty is seen rather than read.
What filters the page
Two things, and only two:
The current activity. As everywhere in Sentinel, the dashboard shows one activity at a time. Switching activity recomposes the whole page, with one exception: public reports carry no activity, so their numbers stay the same.
The date range. There is no other filter: no filter by trail, by assignee or by project. For that, go through the Tasks or Structures sections and their filters.
Who has access
The dashboard falls under the Dashboard right in the activity manager, at the read level. Without it the entry doesn't appear in the menu and the app opens elsewhere. The cost block additionally requires the Costs right.
It's a right to grant broadly: a dashboard allows no changes, and a crew that sees its numbers steers better than a crew that merely endures them.
Three pieces of advice
Look at the shape before the number. A number on its own says nothing. "Open tasks: 68" means nothing; "68 and the line has been rising for ten days" means it's time for a work day.
Lengthen the range to judge, shorten it to act. A month shows a trend, seven days show what just happened. They read differently.
Finish with the trail table. The balances say there's a problem; the table says where, and its first trail is the one you can act on come Monday.