Data Discovery

With Version 3.2 we introduce Data Discovery. This add-on to our Swarm Control Center enables Partners and Customers to easily show, check and validate Swarm generated data.

What is Data Discovery?

  • A tool to show, check and validate Swarm generated data

  • A simple overview for Partners and technical driven End-Users

What is Data Discovery NOT?

  • Fully qualified Dashboard Solution

  • Data aggregation software

Whom is Data Discovery designed for?

  • Swarm Integration Partners

  • Engineering Offices providing Traffic Data

How does Data Discovery work?

Data Discovery sends generated events to the Azure IoT Hub, where they are collected (via MQTT) and stored into a database (mssql). Via API the data is provided to the Swarm Control Center and can be used for creating a widget-based canvas.

Sample Canvas with Widgets

In case you are using your own MQTT broker, you cannot use Data Discovery. You may only use one option.

First steps

After you did the configuration of the Swarm Perception Software, navigate to Data Discovery in the lefthand menu. The very first time, you will get the message showed below:

Use the button "Add Widget" or click on "Widgets" in the top menu to create widgets for your canvas.

Choose your metric

You can choose between CL (Counting Line) Count and various ROI (Region of Interest) metrics.

Choose your data

This part is optional. Besides the metric you can show vehicles classes and subclasses, the name or ID of any CL or ROI, CL Directions, Timestamps and more.

Define Time constrains

You might be interessted in a specific timeframe and a grouping (like hourly). This part is quite intuitive, just select the time, which data you want to see and give it some grouping. E.g. Select data for Today split by Hour will show data grouped by hour for the current day.

Besides pre-definded times, like "All time", "Today", "This week", "Last 7 days", etc. you may also define your own time intervall, choosing "Custom".

Define Output

We give you 6 options to present the data:


Filtering can be done on the same fields, as in "Choose your data". Let's say, you only want to count bicycles. This is the option to go with:

Especially, if you have multiple Swarm Perception devices running, you want to filter your widgets for each devices or even for each configured CL or ROI. Remember: You may use multiple camera streams per device.


All the data can be set into order. It's even possible to order by multiple fields and give them a hierarchy, dragging and dropping them.

ASC: Ascending. Sort alphabetical from A- to Z or numerical from 1 to 9. DESC: Just the other way as ASC. NONE: No order involved with this field.

Last steps (finishing)

At the bottom of the page you will find a preview of the widget you just created. Here an example with these parameters:

  • Metric: CL Count

  • Data: CL Device Name, CL Class, CL Direction

  • Time: Today, split by Hour

  • Output: Table

  • Filter: CL Device Name

  • Order: CL Timestamp ASC, CL Direction ASC

As soon as you are happy with the widget you created, click on "Add to Canvas". A Pop-Up will ask you about a name of the widget. You will be automatically redirected to the Canvas view and see your widget there. With drag and drop it's possible to move the widget around.

Widgets will automatically be fitted on the grid, targeting the left top. In addition they do have a minimum size.

Can I export the data?

Absolutely! You may export the data of every widget, direct from the canvas view! The exported data will be in CSV format and can be imported into Microsoft Excel as well, as processed with any tool, which is able to work with CSV files.