Active alerts

Managing your active alerts is the most important workflow to add into your schedule and/or processes. You can make use of e-mail alert notifications to receive instantaneous notifications when alert conditions are triggered, and/or you can use the Alerts screen in Centrix to see all the currently active (triggered) alerts and interrogate them.

Standard alert features

Centrix supports a number of preset alerts for various known rail faults/conditions. However, each alert shares the following standard profile;

Alert name

A unique name for the alert, this will appear in the e-mail subject of any notifications.

E-mail

List of e-mail addresses that should receive e-mail notifications.
One e-mail address per line.

Tip
Tip: Use labels to group e-mail addresses into logical groups, rather than using the e-mail field. Tip: Even when using labels for alerts, it is useful to use this when 'trialling' a new alert that you don't want e-mailing to other team members.

Subscribed?

E-mail notifications will only be sent if this is ticked .

Suppression period

Use this to stop Centrix sending notifications for repeated alerts.
Repeated alerts within the configured period will not trigger e-mail notifications, but will still appear in the alert history in Centrix.

Priority

Use this as desired to allow you to make a decision on prioritising active alerts.

Tip
Tip: Use High/Critical for alerts on critical assets. Tip: You can setup multiple alerts for track circuits/points with High/Low thresholds.

Labels

Labels are used to group alerts into their associated teams. They can also be used for informational purposes.
Labels can have e-mail addresses attached to them which will define who should receive e-mail notifications for the alert they are attached to.

Alert listing

The alert listing (access through the Alerts menu link) shows all configured alerts within your user group.

Tip

For managing active alerts, you should select the Active Only filter at the top of the listing.

To view only the alerts for your team, filter the listing by Label using the Filter by Label filter at the top of the listing. You can select more than one label if desired.

Tip: If you have a lot of active alerts and not currently in a position to clear out older ones, sort the listing by Last Triggered (descending) to see the most recent alerts first. Click on the table column header to change the sort.

Each alert will have 1 of 3 possible statuses;

Clear

Alert has not triggered since it was last cleared or created

Active

Alert has triggered since it was last cleared or created

Active - Returned Normal

Alert has triggered since it was last cleared or created, but it has since returned back outside the threshold conditions

Info
This feature only applies to the Points alarm type.

Start your investigation

Double click on any alert to see the history of alert raises (triggers). Here, the Message will provide you with details about why the alert triggered.

Click the button in the Actions column to open the graphical replay panel at the point in time the alert triggered.

See the section on Graphical replay for further details.

You can save a Bookmark for any alert raise by clicking the button in the Actions column. This is useful if you want to quickly return to the fault in the future, or you want to share it with a colleague. More on Bookmarks later.

Clearing alerts

Once you are done dealing with an alert, you should clear it so it no longer appears in the listing of active alerts.

You can clear a single alert by clicking the button in the Actions column of the alerts listing.

You can clear one or more alerts using the Bulk Clear action found in the bulk actions drop down at the bottom of the listing. First select all the alerts you want to clear, then click this action.

Graphical replay ("Playback")

The graphical replay panel is the heart of the Centrix system and provides a configurable panel for viewing various types of data and Centrix features on a single screen.

There are a few methods to access the graphical replay panel, depending on your starting point;

From the dashboard

When your first enter Centrix, you are presented with a simple screen that allows you to select a Site and a Date/Time to view data from.

From the sites listing

On the Sites listing you can double click on any site to open the replay panel. The panel will open at the current date and time.

From an alert raise

When viewing the raises of an alert, you can click the button to open the replay panel at the date/time the alert triggered.

From a bookmark

You can open the replay panel at the date and time of a saved bookmark by clicking the button on the bookmarks listing.

Replaying data

The top bar of the replay panel contains some playback controls that allow you to control the automated playback of event data;


Jump back 1 day	Jump back 1 hour	Play backwards*	Pause replay	Play*	Jump forward 1 hour	Jump forward 2 hours	Select the play mode

* Play (forwards or backwards) will automatically replay the data in the mode selected in the drop down (real-time, real-time*10 or event step)

Play modes

Real-time	Plays the events through in real-time as they occurred
Real-time * 10	Plays the events through at 10x speed as they occurred
Event step	Plays the events through one by one at a fixed interval

Current date/time

The current date/time is shown in the centre of the top bar. When the data is playing, you will see the time ticking forwards/backwards depending on the selected play mode.

Tip
You can click on the date/time to manually select a date/time to jump to.

Each panel has a black bar in the centre which indicates the current date/time in the data set. The item before the line is the last item that occurred before the current date/time, the item after the line is the next item that will occur once the date/time reaches the time of that item.

Example;

Current date/time:

Digital events panel:

Manually stepping through events

You can use the scrollbar on any of the event panels to scroll through the data;

Click the up/down arrows to move one event at a time.
Click within the scrollbar to quickly move up/down the events.
Click and drag the scroller to move to a specific position in the event listing.
Scroll with your mouse wheel to move through the events.

You can also use the arrow keys and page up/page down functions on your keyboard to move through events. Make sure you click within the box first to select it.

Map

The map provides a facility to visualise the raw data in a human understandable way. Centrix provides a number of widely used railway assets that can be attached to relay functions and displayed on the map based on relay logic.

Any channel monitored in Centrix can be displayed on the map, even where there isn't a specific railway asset type for it. You can display relay functions as on/off or up/dn indicators which various colour configurations. Analogue data can be visualised as a numeric value or as a gauge.

The current status of the map indicators will be based on the current date/time selected. If the current date/time is now, or in the future, then the map should show the current live status of all assets on the site (minus any lag between the data logger and Centrix).

Tip
If you want to use the replay panel as a "live display" of the site, you should select a date/time in the future, then hit the Play button. This will ensure the current time is ahead of all events, so the most recent events will always be represented on the map as they are received from the data logger. This could be used in an operations environment as a subsidiary system.

Hover you mouse cursor over any item on the map to display the functions that are attached to it, and their current status.

Widgets and features

This section describes each widget panel on the replay panel and what features they provide.

Adding/removing widget panels

Widgets can be added and removed from the replay panel by clicking the Configure button in the top right of the panel. Tick or untick widgets to add/remove, then click Apply. Clicking Apply and save will save the configuration for future visits to that site.

Digital Events

The digital events widget shows all the relay function events received from the data logger, displayed as;

Time (including milliseconds)	Function	State (boolean)*
08:33:42.030	AK TPR	UP

Filtering channels

By default, the listing will show events from all the functions available on the site. You can filter by a specific set of channels by clicking the button in the bottom of the widget.

You can click and drag around the channels you want to filter on the map.
Or, you can manually select/de-select the channels in the listing.

Click Apply Filter to apply the filter.

Show events with time context

By default, the events are displayed in a simple listing, without any context of time between various sets of events without looking at the time of the events.

You can enable time spaced mode on the event listing by clicking the button at the bottom of the widget. Gaps will now be shown between events, relative to the amount of time there was between them. This allows you to easily see which events occurred around the same time, or where there was a gap between events.

Locate a function on the map

If you want to find a function on the map from an event, select the event in the listing and click the button. The map will centre on the object that represents that function and will highlight it for a short period of time.

This is particularly useful for larger sites, or on sites you are not familiar with.

Bookmarks

The bookmarks listing shows the bookmarks that have been created on a site in a time series, similar to the digital event listing.

You can navigate to the date/time of a bookmark by scrolling through the listing, just like you would for the digital event listing.

New bookmarks can be created by clicking the button at the bottom of the widget. The current playback date/time will be used for the bookmark date/time, though you can change this manually if desired.

Alerts

This alerts widget displays a time series of all alert raises for the site. You can scroll through them to jump to instances of an alert raise.

Tip
The alert widget is a useful widget to apply and save on a site.

Logic Analyser

The logic analyser widget provides a way to view the digital logic on a time series, allowing you to compare the pick/drop times of a set of relays.

The current time is shown by the vertical black bar (like the horizontal bar on the event list, bookmarks and alerts list).

You can amend the functions that are shown on the analyser by clicking the button at the bottom of the panel. Select channels in the available or active listings and click the <-- and --> buttons to add/remove channels.

Change the displayed time window by adjusting the Scale using the slider at the bottom of the panel.

Maintenance Windows

Maintenance windows are used to schedule maintenance (either one-off or regular) periods. They are used to suppress any alerts that may occur during maintenance, and to ensure that data collected during maintenance is not used as the inout to any automatic alarm threshold training (more on that later).

If any alerts firing during a schedule maintenance period, they will be suppressed in the system with the reason "maintenance", so you can still see them in the alert history, but e-mail notifications won't be sent.

The widget will show all the configured maintenance windows for the site; past, present and future.

Create a maintenance window

Click the button at the bottom of the panel.

Enter the starting date/time for the maintenance, and select how long the maintenance should last for.

You can put the whole site into maintenance mode by selecting the Suppress all channels option.
Or, you can select individual channels to put into maintenance by selecting them in the Available channels listing and clicking the button.
You can also select channels on the map by drawing a box around the assets you want to put into maintenance mode, then click the Add map selection button.

If the maintenance is one-off, then simply enter a reason (optional) and click Save.

If the maintenance is regular (e.g. every week, every day), then click the Repeat checkbox and select the interval of the repeat (e.g. 1 day for every day, 7 days for once per week).

Tip
Create a repeating nightly maintenance window to automatically suppress permanent-way maintenance generated alerts during non-operational hours.

Bookmarks

Bookmarks provide a method of saving and labelling a point in time for a site. A common use case may be to correlate a reported incident with the associated data in Centrix.

For convenience, Centrix provides the following three methods for creating bookmarks;

From the bookmarks listing

Select Bookmarks from the main menu and click the button at the bottom of the listing. You will need to provide a name for the bookmark, and select a site to attach it to. By default, it will select the current/date time as the bookmarked time, but you can change this to your desired time.

Enter comments to store more detailed information about the bookmark - useful if you want to share some information for colleagues.

Click Save Changes.

From the playback panel

This method is useful when you're in the graphical replay panel and have located an incident or event that you wish to bookmark.

Make sure the Bookmarks widget is enabled by clicking Configure, tick Bookmarks and click Apply.

In the bookmarks widget panel, click the button. The current playback date/time will be set as the bookmark time, but you can manually adjust this if desired. Enter the bookmark name and comments then hit Save.

From the alert raise listing

If you want to create a bookmark for the date/time of a Centrix raised alert, you can do this from the Alert Raises listing in the Alerts screen.

Double click the alert in the alerts listing to open the raise history. On the raise you want to bookmark, click the button. This will take you to the Create Bookmark panel, with the date/time pre-filled as the time the alert was raised.

Virtual channels

Centrix provides the ability to create virtual channels based on simple boolean logic to combine multiple functions into a single digital output. Centrix also provides a timer function to time between two functions which will provide an analogue (time duration) output value each time the condition is met.

The details for creating virtual channels is well documented here; Creating Virtual Channels.

Useful examples;

Combine all lamp checking relays into a single function for simpler alarming.
Code Block
Lamps Out = not ([dip:214 ECPR] and [dip:215 ECPR] and [dip:216 ECPR])

Time between points detection relays to calculate swing times.

Code Block
800 N-R Time = time_between([dip:800 NWKR]DN, [dip:800 RWKR]UP, 10 secs)

Time between level crossing barrier relays to calculate barrier raise and lower times.
Code Block
YN Raise Time = time_between([dip:(YN) MR ZPR]UP, [dip:(YN) MR ZPR]DN, 20 secs)
Time between level crossing barrier close to barrier open to calculate road closure times.
Time between level crossing strike-in to track circuit occupation to calculate warning times.
Time between track circuit occupied to unoccupied at station platforms to calculate dwell times.
Time between successive track circuit occupations to calculate train speed.

Reporting

Points RCM

Centrix provides remote condition monitoring facilities specifically designed for points machines. To monitor points machines, you will need to make use of either the SA380TX or SA380TX-L data loggers which provide analogue inputs.

Once configured, the data logger will send Captures to Centrix each time a points movement is detected. This data is bundled into a single high-resolution waveform at a fixed interval, usually 10 milliseconds.

Background

The general profile of the current draw from a points machine, looks like this;

(y-axis is current draw in amps, x-axis is duration in milliseconds)

Image Removed

When the profile moves away from the baseline, it can indicate that maintenance is required and/or a fault is developing.

Image Removed

Certain aspects of the profile can indicate specific fault conditions. Whilst Centrix doesn't currently have a facility to provide any insight into the type of fault, by developing a catalogue of known fault traces and their relevant traces, engineers can use that to inform maintenance.

Inspecting captured traces

Centrix provides a simple facility for viewing the traces and comparing them on a graph like above. This is accessed through the Captures menu link.

First select a site with points RCM, then select the points asset you want to inspect data for.

Centrix will show a list of all the most recent captures for the selected asset, showing; date/time of the capture, direction of the movement and duration of the movement.

Image Removed

Double click on any capture in the list to view the profile on a graph. Repeat this for multiple captures to compare.

You can add captures from different assets or sites if desired, just select a different asset and double click on the captures you want to add. The graph in the right hand panel will keep all the selected traces until you clear it or reload the page.

Tip
You can export the graph as an image (jpeg or png) using the Actions drop down at the top right. Useful if you want to send it by e-mail, or save a copy offline.

Setting baselines

To support the analysis of a potential fault trace, it is useful to be able to compare any selected trace against a known "good" trace, or baseline.

In the capture listing you can tag a trace as a baseline by clicking the Image Removed button in the Actions column. Centrix provides two baseline types; maintenance baseline and major baseline. You can use either, as desired.

You can find baseline traces by select the tag in the tag filter above the capture listing.

Image Removed

Tagging faults

Similar to setting baselines, you can also apply a Fault tag to known fault traces. This will help towards creating a catalog of known faults and what caused them. You can add comments to any trace to provide additional information about the fault (or baseline) by clicking on the Image Removed button in the Actions column.

You can find fault traces by selecting the tag in the filter list as described above for baselines.

Configuring alerts

Centrix has an alert type specifically designed for monitoring points machine captures. It provides the ability to create thresholds on the following features of a capture trace:

Average	Average current draw taken from all the samples in the trace
Duration	Duration of the trace (i.e. swing time) to the millisecond
Maximum	Maximum current draw (this will be skewed by the in-rush for motor current traces, more on this later)
Total	Total current draw, calculated as a sum of all data points in the trace (there is a data point every 10ms)

Whole trace thresholds

When setting up a threshold for points traces, we recommend only configuring thresholds for Average and Duration.

Maximum is only going to detect changes in the in-rush, which does not generally give indication of any common fault types. Total is likely to cause false alerts, or duplicate alerts that would always be picked up by the duration and/or average thresholds anyway.

Episodes

In many cases, you want to set a threshold for a certain region within the trace, rather than the whole trace. This is particularly true for alerting on points machine current traces, where you want to threshold on "maximum current draw" but need to exclude the in-rush.

Drawio

autoSize	true
border	true
fitWindow	false
diagramName	EpisodeThresholding.xml
grid	false
width	880
pageId	7538776
height	439

Centrix performs a matching algorithm on every trace that is processed to best fit the configured episode(s) to the trace, and uses that to work out the region that should be checked against the alarm thresholds.

When setting up thresholds for the above episode, we recommend only configuring a threshold for Maximum.

Tip
Our points alarm recommendation: For each switch end and direction; create an Average and Duration threshold using the whole trace and create a Maximum threshold using the in-rush excluded episode.

Track circuit RCM

Centrix provides a suite of reports that have multiple uses. They are accessed through the Reports main menu.

Report types

State transition counts	Use this to chart the number of relay transitions over a given date range One or more channels can be compared on the same chart Counts can be grouped by a period (e.g. By day, month, year)
Raw digital events	Use this to export digital event data for offline analysis Exports selected channel(s) and date range into a CSV file
Analogue histogram	Displays the range and frequency of analogue values on a histogram Shows the average value for the given period Used to find the operating range of a given channel
Sequence timing	Takes a given sequence of digital functions to count and time the sequence Add any number of events into the sequence to prove the sequence Select any event as the start and end events so you can add events before/after the timed events to prove a sequence Whilst providing an output of times, it is equally as good at providing the count of sequence matches Provides a full listing of every sequence match, for further analysis
Points swing count	Designed specifically to display number of point operations for a given site and date range Can be used to feed into maintenance schedule Displays number of invalid operations where detection was lost and gained in the same direction whilst the track circuit was occupied Find instances of track flicks Find instances of trailing run throughs
Points alarm thresholds	Displays all the currently configured points alarm thresholds Useful to check and compare all configured thresholds across all assets Useful to indicate where sites don't have alerts configured

Examples

Below are some example operational questions that the reports can answer, but there are many more;

How many times do barriers operate at a given level crossing?

Report	State Transition Counts
Conditions	Select the level crossing site and date range you want to report data for You can group the counts by a specified period (e.g. By day, month, etc) Limit the transitions only to Dn->Up to avoid doubling the overall count. Add the (UP) KR channel to the channel list
Output	Image Added
Analysis	Barriers operate approximately 1,800 times per month There were no major deviations from this

What is the average arrival time at a given AHB crossing? What the deviation from that average?

Report	Analogue Histogram
Conditions	Select the level crossing site and date range you want to report for Select the arrival time channel (must be configured as a virtual channel) Select the number of histogram bins you want to distribute the data over (leave default) You can exclude outliers by limiting the range (leave default)
Output	Image Added
Analysis	All arrival time between 26 and 40 seconds There were 497 trains within the selected range Average arrival time was ~30 seconds. There were ~40 circumstances (~10%) of times greater than 1 standard deviation from the mean

What is the road closure time at a given crossing?

Report	Sequence Timing
Conditions	Select level crossing site and date range you want to report for Give the report a title "Level crossing name - road closures" Create the sequence of events that proves a road closure, e.g. (UP) KR - DN - Start (UP) KR - UP - End Enter the designed time (optional)
Output	Image AddedImage Added
Analysis	Majority of road closures between 70 and 200 seconds Longest road closure was 436 seconds 80% of closures within 190 seconds 50% of closures within 140 seconds There were 555 closures within the selected range

What is the usage (number of operations) of all the points assets for a given site?

Report	Points Swing Count
Conditions	Select site and date range you want to report for Select the "Points" asset group
Output	Image Added Image Added
Analysis	600 points is the busiest asset, operating 2,331 times last month 809 points is the least used asset, operating just 187 times last month There were 3 circumstances of invalid movements

Points RCM

Centrix provides remote condition monitoring facilities specifically designed for points machines. To monitor points machines, you will need to make use of either the SA380TX or SA380TX-L data loggers which provide analogue inputs.

Once configured, the data logger will send Captures to Centrix each time a points movement is detected. This data is bundled into a single high-resolution waveform at a fixed interval, usually 10 milliseconds.