Check back here regularly for the latest features
KPI Rules - trigger on change
KPI Rules can now be triggered on the change of a KPI value. This allow you to send alerts when Uptime has changed, not just when it breaches a threshold:
We've update our chart module to deliver a set of minor improvements around labelling, tooltips and improved layout, which makes more room for your data and everything a bit clearer.
Setting time in KPI Rules
We've improved KPI rules so that you can set the period (the time over which the KPI is run) and the cadence (how often we run the KPI) separately. This allows you to, for instance, trigger an alarm on a rolling average of temperature: test every 15 minutes whether the average temperature over the last hour was greater than 5.
Actions can now be triggered by KPIs.
Our existing Rules system allows you to create action based on individual device behaviour. KPI rules mean you can now apply this to any KPI, such as above, where we've setup a Slack notification which will fire when any region's uptime falls below 90%.
For more info, see rules.
The List view now supports showing a hierarchy of devices, using the group defined in a KPI. This makes it easy to visuale performance by i.e. vendor, gateway, customer etc.
For more info see List View.
Improved tabular formatting
7th January 2020
We've improved the way numbers are formatted in the tabular view and also the way they are exported to CSV.
Markers on Line charts
You can now switch off markers on your line charts to make the view cleaner:
We've had dark mode for a while, but now we can take the setting from your operating system setting on Mac or Windows - just head to User Settings and choose System under Theme.
We've improved the layout of printing dashboards, which also allows you to save the dashboard as a PDF report.
Time selection in Dashboard
We've updated the time selector in the dashboard to allow more flexible time ranges. As well as the current Rolling days option, you can now select complete days, weeks or months, or simply choose a start and end date from a calendar.
We'll migrate the Periods options to the KPI definition soon too, so you'll be able to define KPIs for, say, the last 3 months.
We've added direct integrations for:
EMnify - an IoT connectivity provider
1NCE - an IoT connectivity provider
Eurotech - an IoT Integration Platform
Multi-line map tooltips
You can now add multi-line tooltips to the map, to display as many fields as you like.
Enumerated Calculated Properties
You can now add enumerations to Calculated Properties to convert i.e. fault codes to fault strings.
List export now supports KPIs
We fixed an issue where exporting a list of devices didn't export the KPI if used. We also fixed a sorting bug in the list.
You can now easily invite everyone from your company, but inviting the domain. Don't worry - we won't email them all, but if they log in via Google, we'll automatically put them in the company account, rather than creating a new account for them.
API for KPI results
You can now access the results of any KPI via our API. This allows you to embed our analysis in your application or dashboard with ease. For more info, see Querying data via our API.
In our Rules module, you can now send out notifications via SMS.
We've extended the boundary functioning of our charts to include absolute values. This makes it easy to tell, at a glance, whether your results are within or without a desired range.
Folding Tabular View
KPI results can now be viewed in a folding table. By clicking the column header, you can dynamically group matching rows to help quickly navigate and spot trends in larger datasets.
You can use DevicePilot's webhook actions to connect to any REST API. You can also now fire those actions manually in the View page. Simply click the Action button on the toolbar to fire the webhook manually on the selected device(s).
Further information on webhooks can be found here.
Calculated properties allow you to derive new telemetry from your existing data. Whether it be a simple unit or type conversion, or a more complex differentiation from energy to power, they allow you to create enriched data for your analysis.
Time range selection
You can now choose your time range in the View screen from any point in your history.
All the other views will work just as expected, except that "now" won't be "now", it will be the point in time that you select.
This is particularly useful to analyse a performance issue that happened a few days ago.
Drag and drop Floorplans
In the floorplan widget, you can now arrange your devices by dragging and dropping the devices. Once you go into Dashboard edit mode, you'll see a panel on the floorplan with the available devices. Just drag them onto the floorplan, and you are good to go.
Map drill down
You can now select devices in the map widget on the dashboard and drill down to View, keeping the selected device as you do.
We've added our first piece of machine learning into the analysis engine - you can export much more to come. We start with a forecast function on device telemetry - we use a machine learning forecast algorithm to predict the future values of the data:
This also opens up the possibility of anomaly detection - where the data didn't match the forecast.
You can now choose a time range to apply to all dashboard widgets. The default is to use the time range specified in the KPI, but now you can override it globally.
We have renamed monitors to rules, and made the process of building them much more straightforward.
More details available here: How do I perform actions on my data?
We now have a direct, zero-code integration to Disruptive Technologies sensors. With this integration, you can get connected to your sensor data in seconds - for more information see How do I connect to Disruptive Technologies.
KPI Pie Charts
KPIs can now be displayed as pie charts, if they are grouped by a property.
You use the legend to remove values you don't want see, save the KPI and we'll remember the setting, both in Cohort and on the Dashboard.
In the map, we now display a legend when you are colouring the pins by a value. We've also added a more standard blue -> green -> red colour scheme (and the reverse).
Splitting Chart series
A new feature on the History Chart allows you separate a single chart series into multiple by splitting on a property. The use case we built it for was a wifi test device measuring throughput on two different SSIDs and wanting to see those separately.
DevicePilot is now responsive on mobile - this means you'll get a view of your dashboard in a clean, mobile-friendly view.
We've just worked on dashboards for now. Let us know which area you think we should work on next.
Along with this, we have a few other small enhancements to make life easier:
- View Details pane now includes updated time per-property
- Properties table in Settings shows hidden properties in grey
KPIs now store time range
In the Cohort page, we now save the time range for KPIs as part of its definition. If you've used the KPI in a dashboard widget, then we've set the time to that, otherwise it's the default 24 hours. For now, you can still override that time in the widget.
We've also fixed a number of minor issues with selecting time, devices and properties in the View page - all to make it easier.
DoD, MoM, QoQ and YoY comparisons
You can now add a comparison to a previous period in any KPI that is grouped by time. You can compare Day on Day, Month on Month, Quarter on Quarter and Year on Year.
You have three ways to display this:
- Absolute values, side by side
- Absolute difference
- Percentage difference
One for the advanced users here. In the Cohort page, you can now use an expression instead of just a property in your metric definition. The syntax supports simple expressions with mathematical operators.
You can now show single value dashboard widgets as a gauge.
This is particularly suitable for values with expected ranges - you can specify the min and max and quickly visualise whether the value is expected. For percentage KPIs, we also use the RAG colours.
KPIs in Map Widgets
You can can setup a map widget with your KPI on your dashboard. Open the widget settings and choose the KPI.
In the View screen, when using our new KPI visualisation feature, you can toggle the RAG (Red Amber Green) bands off so you can focus on the devices your are interested in. You can also simply toggle off the devices that are either 100% or 0% working, since these also may be of less interest.
You can configure your RAG levels in the Cohort page, and we'll use then in the dashboard widgets and in this new View legend
Google Cloud IoT Core Integration
We can now connect directly to your Google Cloud IoT Core instance, with no code for you to write. The process involves you creating a key in Google Cloud, and we take you through the process step-by-step. For more details, see How do I connect to Google Cloud IoT Core?
When analysing a KPI in view, we now display a timeline view showing the history of the KPI under analysis, just for the selected device. For more details, see View Timeline.
We have also updated our Google login buttons to match their new, strict brand guidelines.
Direct integration with Azure IoT Hub
You can now stream data directly from Azure IoT Hub, with an extremely simple setup. Read more about it here: How do I connect to Azure IoT Hub.
We'll continue to build more direct integrations - next up Google Cloud IoT Core.
KPI drilldown in View
From today, you can now see your KPIs down to the device level in View. This enables you to quickly see the worst performing devices and analyse those ones first.
You can select the KPI to analyse from the picker, or alternatively, you can drill down from a dashboard widget to open this view.
See Visualising your data for more details.
Some of our customers have data coming from in-building sensors. In those cases, even if the Google maps have good indoor coverage, they asked us to be able to draw data on top of a floorplan image.
So now, in the dashboard, you can add and configure this widget to show your devices like this:
Lock Map widget bounds
A small thing, but important to make nice dashboards. The default behaviour of our map widget is to zoom to fit the extents of the data each time we update. Some of you told us that you would prefer to be able to set the boundaries of the map and have them fixed.
Now, when in edit mode on the Dashboard, you'll see that you can lock the view. This will remove the zoom icons and mean your view will stay as it is.
Map History Mode
This is a feature for those with devices that move around. In the map, you can toggle between showing:
- Estate Mode - all devices at their current location
- History Mode - the selected device at all of its previous locations in the time window
The big green dot is the latest location of the device.