What a dashboard does: monitoring, decisions and trends on one screen

A Break at 6:40 in the Morning. Picture this. It's six forty on a cold Tuesday morning, and a water main breaks under a busy downtown street. The call center starts getting reports of low pressure. The field supervisor wants to know which crews are free. The operations manager wants to know how many customers are without water. The city manager wants to know if a school is affected. Four people are asking four questions, and they're all phoning the same two analysts. Now imagine each of them could open one screen and see the break on a map, the number of affected customers, the crews assigned, and the reports still coming in. That shared screen is a dashboard, and building good ones is what this course is about.

What You Will Learn. By the end of this lesson you'll be able to do three things. First, you'll name the three jobs a dashboard does: monitoring what's happening now, supporting decisions, and showing trends over time. Second, you'll look at a screen, like a storm response or service request dashboard, and say which of those jobs it's actually doing, and whether it's doing that job well. Third, you'll trace the path data takes in ArcGIS Online, from a hosted layer, through a web map, onto the dashboard itself. That path matters, because most dashboard problems start upstream in the data or the map, not in the dashboard.

What a Dashboard Is. A dashboard is a single screen that shows several views of the same data side by side, so you can understand a situation at a glance. Think of the dashboard in your car. You don't read a report to learn your speed or fuel level. You glance, and you know. In ArcGIS Online, the tool for this is ArcGIS Dashboards. You arrange elements on one page: a map, big number tiles called indicators, charts, lists and filters. They all read from the same data, and when you click a break in the list, the map can zoom to it and the charts can filter to match. When the data updates, the screen updates too.

Job One: Monitoring. The first job is monitoring. A monitoring dashboard answers one question: what's happening right now? Think of a utility outage screen on the wall of an operations center. It shows how many outages are open, how many customers are without power, and where they are on the map. The dashboard refreshes on a schedule, for example every minute, so nobody has to press reload. Traffic management centers use the same idea for incidents, and emergency operations centers use it during storms. The key feature is freshness. A monitoring dashboard with data that's six hours old isn't monitoring anything. It's just a nice picture of the past.

Job Two: Supporting Decisions. The second job is supporting decisions. Here the screen isn't just telling you what's happening. It's giving someone what they need to act. Follow the flow for our water main break. A break is reported and appears on the dashboard. The supervisor filters to the north district, sees which crews are free and how far away they are, picks the nearest one, and dispatches it. Every step used the dashboard. So what makes this work? The numbers are tied to places. A count of twelve open service requests means little on its own. Twelve requests clustered on three blocks around the break tells you exactly where to send people.

Job Three: Showing Trends. The third job is showing trends. Instead of right now, it asks whether things are getting better or worse over time. Here are water main breaks for five cold-season months in one city. January stands out with thirty-eight breaks, almost three times November. Cold ground shifts and old cast iron pipes crack. Drawing the monthly average of twenty-four point two makes the pattern obvious: the middle of winter is well above normal. That one chart gives a manager a clear case for pre-positioning crews and parts in December next year. Trend dashboards are usually read by managers and executives, and they guide budgets and planning rather than tonight's dispatch.

Worked Example: Two Indicator Numbers. Let's make an indicator concrete. An indicator is one of those big number tiles, and behind it is a simple calculation over your data. Don't worry about the code itself: no coding is required in this course, and this is only here to show the logic. Here are five power outage records, each with a status and a count of customers affected. First we keep only the rows whose status is Open. Then we count them, and we add up their customers. The result is three open outages and two thousand three hundred eighty customers without power. ArcGIS Dashboards does exactly this for you when you set an indicator to count features, or to sum a field, with a filter on status. Now you know what the tile really means.

How the Pieces Connect. Here's how data reaches the screen in ArcGIS Online. Reports come in from a field app or a three one one service request system and land in a hosted feature layer. That's a table of records, each with a location, stored in ArcGIS Online. A web map then displays that layer with symbols, pop-ups and visibility settings. On the dashboard, the map element shows the web map, and charts, indicators and lists read the same layers. People view it on a wall screen or a phone. This chain matters when something looks wrong. If a date shows as a meaningless number, fix the field in the layer. If breaks are hard to see, fix the symbols in the web map. The dashboard only shows what it's given.

Four Questions Before You Build. Before you build anything, answer four quick questions. Who looks at this screen? A crew supervisor, a call center agent and a council member need very different things. What question do they bring? For the supervisor it might be, where are the open breaks and who's free? How fresh must the data be? Minutes for a storm response, a day for a monthly trend. And what will they do next? If the answer is nothing, the element probably doesn't belong. Get these four answers written down first, and the layout almost designs itself, because each element on the screen has a clear reason to be there.

Common Mistake: The Everything Screen. The most common mistake is the everything screen. Someone gets access to ArcGIS Dashboards and adds every chart they can think of: breaks by pipe material, breaks by year since nineteen ninety, live crew locations, budget totals, all on one page. The result is small text nobody can read from across the room, and a screen that answers no one's question well. Live monitoring and five-year trends also refresh at very different speeds, so mixing them confuses people. The fix is simple. Pick one audience and one main job per dashboard. If executives want trends and crews want live status, build two dashboards from the same data.

Recap. Let's recap. A dashboard is one screen with several connected views of the same data, and it does one of three jobs: monitoring what's happening now, supporting a decision someone needs to make, or showing a trend over time. Each number tile is a simple calculation over your data, like counting open outages or summing affected customers. And the data travels a clear path: a hosted feature layer, then a web map, then the dashboard. When something looks wrong on screen, fix it at the right level. In the next lessons we'll match dashboard types to their audiences and get that data ready.

Your Exercise. Here's your task before the next lesson. Pick one scenario you know well: water main breaks, power outages, traffic incidents, service requests or storm response. Write down who will look at the dashboard and the one question they bring to it. Then decide its main job: monitoring, supporting a decision, or showing a trend. Finally, list four elements you'd put on the screen, such as a map, an indicator, a chart or a list, and write one sentence for each explaining why it belongs. If you can't justify an element, cross it out. Keep this sketch, because we'll build on it throughout the course.