Thursday, December 07, 2017

An alternative system for SMRT to consider

INTRODUCTION
I question the decision to invest $195 million in the new signal system for the NS and EW lines.

I find the amount to be excessive. I also dislike the many instances of "signal faults" that caused the frequent disruption in the train services and the collision at Joo Koon station.

Under the old signal system, we did not have the frequent breakdowns that are occuring now.

Could this be caused by bad implementation? Did we rush to implement the new system without proper testing?

More important, will we be able to see the light of day soon. Can we get over the frequent disruptions? How long more do we have to wait to see the improvement in service that was expected from this large investment?

Another question is - do we have an alternative?


AN ALTERNATIVE
I believe that we have an alternative to the Thales CBTC system. I know that they are a reputable system and their system is used in many cities.

But I do not accept that we need to implement an expensive system just because it is used in other cities. The more important question is - is there an alternative?

What is the system expected to achieve?

Take the NS line. It runs for 45 km and has 28 stations. The average distance between the station is 1.6 km.

If we despatch a train every 2 minutes, the average distance between 2 trains is 2 km, based on an average speed of 60 kph. 2 km is a long distance. Even if we reduce the interval of the trains to 1 minute, the average distances is 1 km.

We like the trains to arrive at each station on time. There could be local factors that cause the train to be temporarily behind schedule. The can be rectified by getting the train to increase the speed until they are back on schedule.

There is the challenge to get the train to stop at the exact spot on the platform, so that they train and platform doors can open. I prefer that this be achieved using local signals and sensors, rather than rely on the central system to tell the train exactly where to stop.

There is also the challenge to get the train to slow down or stop to avoid collison with the train or another obstacle ahead. Under normal circumstances, this should not happen, as the trains are all moving at the expected speed and there is a distance of 1 or 2 km between the trains.

But we have to deal with the unexpected emergencies.

COLLISION AVOIDANCE SYSTEM
I prefer each train to have a collision avoidance system. This is already installed in expensive cars. I like a similar system to be installed on each train. Remember, this is to be used only in emergencies. It is not intended to regulate the speed of the train.

CENTRAL SYSTEM
The speed of each train will be regulated by the central control system. This system has the primary goal of instructing the train on the speed that it should travel.

The central system should not be burdened with the task of getting the train to stop at the right spot in the platform (as it can be delegated to the local sensors and systems).

PROTOTYPE
I have developed a prototype of a system that describes how my approach can work. They can be viewed here.

Operation of the system
tklcloud.com/mrt/home.aspx

Signals of the system
tklcloud.com/mrt/page/7

Monitor screen for simulation
tklcloud.com/mrt/monitor.aspx

If you have any feedback, send email to kinlian@gmail.com

No comments:

Blog Archive