A survey on cluster based routing, data aggregation and fault detection techniques for locating real time faults in modern metro train tracks using wireless sensor network

Metro railways are the most widely used mode of public transport in urban areas due to their high speed, capacity and reliability. The current state of art in detecting immediate and long term railway track problems involves both inspectors walking the track lines and train cars instrumented with accelerometers and ultrasonic sensors that are capable of detecting wear of the rail and breakages. But still, accidents do occur. Track maintenance is the primary concern for railways to prevent accidents. Human inspections of the track are subjective and consume a lot of time, labour and power whereas train cars instrumented with sensors will provide the local analysis of the inspected track and are prone to missed failures and false alarms. Condition monitoring using Wireless Sensor Network can provide the integrated view of the entire railway asset and thus can reduce human inspection requirements through automated monitoring. This approach can reduce maintenance time by detecting faults before they escalate and thus, provides a more safe and reliable system. However, to build an efficient railway track fault detection technique using Wireless Sensor Network is a challenging task as sensors are power and memory constrained devices. Therefore, this paper presents the survey on cluster based routing techniques that can be used to reduce communication cost, data aggregation techniques to reduce data redundancy and help to transmit useful information and fault detection techniques to identify the faulty nodes in the network. We have also presented the analysis of these surveyed algorithms based on their approach and parameters like energy consumption and network lifetime which can be used to build an efficient track fault detection algorithm that can provide the track fault location in real time and alert the system with less energy consumption.