THE development of Strukton Rail's Poss remote condition monitoring system started in 1999 in response to the urgent need to improve the availability of points in the Netherlands. It was soon clear that traditional ways of improving point availability would not achieve the required result, so a completely new method was chosen: remote condition monitoring.

The results of the first pilot installation on a busy junction near Amsterdam improved point availability by 35%, which led to the nationwide introduction of Poss point monitoring in the Netherlands.

The hardware has been continuously improved since 1999 resulting in the development of our MicroPoss data logger, with more than 600 units now in use. The software has also been improved over the years mainly in cooperation with our clients.

From the outset we have tried to make the system simple and easy to use. With this in mind, we decided for point monitoring to measure only the motor current of the point drives as this enables the user to find almost 80% of the problems which affect availability. If we attempted to discover more problems the system would become unnecessarily expensive, technically complicated, and far from user-friendly.

All Poss clients use one central cloud server, and each customer pays an annual fee related to the number of connected assets. For this fee we guarantee high system availability and take care of all maintenance activities and system updates.

The success of remote condition monitoring depends highly on the way the system is implemented within the organisation. We help our clients to get the best out of Poss by including a training programme for both the maintenance and fault teams as well as the management. These training sessions take no longer than two hours per group, because the system is so user-friendly.

Poss is designed to be non-intrusive. The motor current of the point drives is measured by current sensors, so there is no galvanic contact to the signalling equipment. Local Poss equipment meets EMC class A specifications as well as EN 55022, EN 61000 and EN 60950, and can be used in electronic interlockings.

With Poss, the current sensors are connected to a local data logger, which is the only hardware the client buys. The data logger is coupled over a GSM, GSM-R, or glassfibre communication network to the central cloud server, which consists of a message broker, data storage, analysis tool and the monitoring applications. The client can access this data via laptop or desktop computer, iPad, or smart phone with an internet connection.

Poss can help to increase line availability by giving an alarm as soon as a point exceeds one of the preset alarm levels which can be set by the client. In up to 40% of cases, the point can be repaired before train delays occur. After the repair, Poss can check whether it was done correctly, which can prevent repetitive faults.

Poss can also optimise the maintenance process. A Poss graph can give the client a good indication of the "health" of a point, with such things as locking, unlocking, obstructions during the swing, and bent blades clearly visible. An analysis tool shows the behaviour of the point over a longer time, making trends and maintenance activities visible. Points can be maintained more efficiently because Poss can determine exactly where maintenance is needed.

The results of remote condition monitoring depend largely on the user. If the user does not like the system because it is too complicated, slow, or has more defects than the points themselves, the results will be poor. A simple, reliable and user-friendly remote condition monitoring system combined with well-trained staff can easily achieve up to a 35% reduction in point failures.

Conventional relay track circuits can be monitored by measuring the current from the track to the relay. Poss software measures a sample of this current when the track is unoccupied every 15 minutes and displays this on a daily and monthly graph. An alarm will be raised when the current level becomes too low.

This monitoring is particularly useful in track circuits with bad trackbed conditions. We are now developing a track circuit monitoring option for audio frequency track circuits.

The same hardware with a different software application can also be used for axle shunt monitoring. In this case, the current to the track relay is measured while the track circuit is occupied. PossAxle shunt monitoring emits an alarm when the rest current to the track relay is too high because of loss of shunt.

Thales axle counters already have a monitoring system available in the signalling equipment rooms, and the data is made visible on the Poss site via an Ethernet connection.

Two Poss monitoring options are available for level crossings. One is a simple solution which only measures the time that the level crossing is open and closed, which is useful on intensively-used lines. All statistical information about the level crossing can be found on the website, such as the percentage of opening or closing per hour or day, and deviation from normal.

The other option is an intelligent video processing solution which can conduct online checks of lamps, barriers, and signs. This solution can also provide useful information in the event of an incident.

Finally, measuring rail temperature is important to prevent buckling when temperatures exceed 25oC, and Poss is able to measure rail temperature and sound an alarm when the temperature exceeds a preset level.

While introducing railway condition monitoring can improve maintenance performance, simply buying a system is not enough. Embedding the system in the organisation is crucial to its success.

Poss applications on rolling stock

MODERN trains have sophisticated on-board diagnosis systems, but there are no specialised staff on each train to interpret the data. Since MicroPoss data loggers are connected over a wireless connection to a central server, it is also possible to install these units in trains, preferably using a GSM-R connection. The on-board diagnosis unit can then be connected via Ethernet to MicroPoss.

Train data is displayed on a website, allowing the trains to be monitored from the depot by specialised staff. This saves time because the status of the train is known long before it enters the depot.

Trackside monitoring systems that check the quality of the passing rolling stock are available.

A WIM measuring point has six laser sensors along a 4.2m section of rail which measures the total circumference of each passing wheel. After processing the data, the train weight per wheel is presented in a graph. The top of the graph shows a schematic overview of the train, and overweight wheels are shown in yellow or red.

Weight in motion measures the following items:

• maximum axleload

• maximum linear vehicle weight

• number of axles and length of train

• speed

• overall static train weight

• dynamic forces acting on superstructure

• locomotive type, number, weight, and position in train

• wagon type, number, position in train, Sofis ID readout

• wagon weight and axleload limit depending on wagon type

• static single wheel loads

• dynamic running behaviour, and

• maximum dynamic load per wheel

For wheel defect detection an additional sensor is placed between the sleeper and the rail. The sensor has the same shape as a sleeper pad and replaces the existing pad. It has a built-in strain gauge and just measures the force between rail and sleeper.

The combination of a laser sensor and sleeper pad can detect the following wheel defects:

• static load per wheel

• Qdyn versus Qstat ratio

• dynamic load per wheel

• axle/bogie/wagon load distribution

• dynamic running behaviour

• running tread defect identification

• running tread and wheel defect type (flat/spalling/out-of-roundness), their number and position

• buffer climbing tendency

• axle or bogie hunting evaluation

• maximum lateral force per wheel, and

• lateral versus vertical force ratio.

The wheel defect sensor can detect the difference between wheels with conventional cast iron brakes and those with silent composite brakes.

For wheel profile measuring, an extra laser sensor is fixed to the rail. The diameter differences (dMdiff) of wheel axles can be determined by evaluating the measurement data from the position of the wheel on the track (distance to the rail head), flange height (Sh) and chord length (SL), and diameter (dM).