Performing a remote diagnostic of a connected product involves a systematic approach to analyze data and events in order to identify the root cause of a problem.
Below are the main steps followed by service users during remote diagnostics, and for each of them in the dedicated section you will find how to configure the DPS to carry them out.
Check the product connectivity.
Analyze the occurred events.
Analyze the collected data.
The remote diagnostic is made by Service users that are technicians, so in the DPS it takes place by accessing the Service view.
Once the technician has identified the machine, it enters the machine details page where he can:
Check Connectivity
In this step, the Service user must check the connectivity of the products and machines they manage.
A bad connectivity may result in loss of meaningful data, which makes remote analysis difficult or even impossible.
Each thing definition has a predefined metric called Connection Status, which is automatically kept updated by the DPS.
When the service user enters the machine details page, it is very important to have evidence of the current connection status, to define whether the displayed data is up-to-date or not.
Into a Thing Details or directly in the Thing List widget, you can display the connection status along the last update timestamp by using a built-in filter.
Even if the object is online, this does not mean that the overall connectivity is good and therefore that the historical data is complete and reliable.
You can leverage the Connection Index insight, which provides a synthetic index about how good the connection is over time, is a percentage value, and a low percentage means a bad connection.
The aggregated value (last 30 days average) of the connection index can be displayed by using the Value widget in the thing definition Overview tab.
In addition, you can define a specific Connectivity tab to provide more details about connection status over time.
Analyze Events
Events are noteworthy situations that are recorded in the cloud when certain conditions occur on the machine.
If defined correctly, events can provide immediate proof of what is happening or happened on the machine.
In the DPS you can leverage different tools to let your service users to analyze events.
Event List: the simple widget providing access to active and historical events. The DPS user can filter and search for specific events.
Clicking on the event will take the technician to the details page where they can see more information about the event, including the technical description and troubleshooting.Events over chart: in the Time Series you can display the events along the data series.
In this way you can help technicians to understand the possible causes of an event (e.g. the temperature started to rise).Event Pareto: in a Bar Chart displaying the number of events grouped by type, you can display the Pareto line, so the technician can focus on the most frequent events (COUNT) or those on which the most time (SUM of durations) has been spent by the machine.
Analyze Data
As a last step, if the connectivity is ok and there are no relevant events, analysis of the data makes it possible to check for anomalous situations.
Firstly, you can use the Value widget to display the latest of aggregated value (e.g. last X days average) of a relevant metric.
If you have configured the Thresholds on the metric, the widget will render the value by using different colors according to the value severity.
Within the Diagnostic tab of the thing definition view, you can place several widgets that can help technicians to analyze data:
Time Series: displays the historical data of one or more metrics. You can also display the minimum and maximum lines.
Multi Metric List: display the historical data of one or more metrics in a table.
Scatter Chart: combines two metrics to identify outstanding data points.
Benchmark: compares the current thing with others with similar characteristics.