• No results found

Automatic Railway Crossing System Based on Load and Camera Sensor and Obstacle Detection

N/A
N/A
Protected

Academic year: 2020

Share "Automatic Railway Crossing System Based on Load and Camera Sensor and Obstacle Detection"

Copied!
6
0
0

Loading.... (view fulltext now)

Full text

(1)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 9, Issue 10, October 2019)

Automatic Railway Crossing System Based on Load and

Camera Sensor and Obstacle Detection

Jaspreet Mehra

1

, Monika Jain

2

, Megha Kamble

3

1,2,3

CSE, LNCT, Bhopal, India

Abstract—Rail transport is the most popular system in

India. With the introduction of cutting edge technology, enhancements are necessary to make it more safe for common people. On this front, this paper proposes improvising for automation of railway gate operation. The arrival and departure of the train near level crossing determines the opening and closing of the level crossing automatically with the help of IoT load sensors and warning signal(alarm or buzzer). During buffer closure time there may be situation where vehicle may be locked on track between crossing gates. It will be detected by camera sensors and intimated to gate closing system through IoT sensors. This could reduce manpower and accidents at level crossing upto maximum extent.

Keywords—Automatic railway crossing, unmanned crossing, Internet of things (IoT), machine learning algorithm

I. INTRODUCTION

Rail transport is most preferred and the least expensive method of transportation used by the general public in India and worldwide. It is preferred the most over the other means of transportation. Human handling of railway gates, lack of synchronization mechanism, ignorance of general public are multiple causes of railway crossing related accidents in routine life. The void caused by the loss of mankind can never be replenished. Therefore, there is an necessity to take some precautionary steps to enhance railway crossing controlling using state-of-the-art technology.

Presently, most of the railway crossings are taken care by gatekeeper ie human being and some are unmanned using various sensor mechanisms. Even literature presents number of techniques for automation of railway gates. Current scenario is of IOT and ML based intelligent and cloud basing controlling proven technology. In general, level crossing gates are operated manually by a gate keeper when the train starts to leave the station, concerned incharge delivers the information to closest gate keeper to get ready. In case of train delays, gates remain closed for long duration causing people waiting indefinitely at crossings. The rate of manual error occurring at these level crossings are high because they are unsafe to perform without actual knowledge about train timetable.

This will lead to ignorantly crossing of tracks by the common man. So these human interventions can be avoided by the automating process.

The Internet of Things (IoT) also called the Internet of objects, refers to a wireless network between objects by embedding short range mobile transceivers into a wide array of additional gadgets and everyday items. IoT enables new form of communication between people and things and among the objects themselves. This is emerging technology and research disciplines that enables the internet to reach out into the real world of physical objects. The Internet of Things (IoT) permits objects to be detected and controlled remotely over existing system structure. Allowing more straight forward combination of the physical world into. computer based frameworks leads to enhanced productivity, precision, effective communication and financial advantage. The IOT application covers smart environments in various domains and found very useful in case of automation of railway gate controlling.

This paper presents a framework by integrating IoT hardware and sensors with machine learning algorithm that would help in preventing rail-transport accidents upto a greater extent. IoT load sensors sense the track for train presence, generating signal to trigger controlling system automatically without human intervention where as there will be short interval in which still obstacle will be there until the track is entirely clear. For this short duration, track images will be clicked by automatic camera. These images are analysed using ML object localization to identify exact location and object detection to classify them and warning time will be given before train approached the crossing gates. So the aim is by controlling the opening and closing of the railway gates without any human assistance and object detection, obstacles on the track are removed upto great extent so as to avoid the accidents.

(2)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 9, Issue 10, October 2019)

59 II. EXISTING WORK

A. Present Scenario

Manual Railway gate controlling is the popular present system. When a train leaves any station, the station master of current station intimates upcoming crossing's gatekeeper about the arrival of the train by a telephonic call. On receiving the manual information, the gatekeeper comes into action to close the the gate manually and the time interval for the said process is approximate time interval which is varying on daily basis. The time interval is approximated by manual calculation by considering the distance traveled by the train from a station to the crossing gate. However, the gate remains closed for a long time even if the train is late for some reasons. If number of trains are crossing, the gates will open only when entire track is free. The natural human tendency triggers the usage of track although crossing is closed leading the track to accident prone area. Unmanageable delays, human handling and unawareness of crowd to hurriedly crossing the track these are the some of the problems to be taken care by automation system. By keeping the limitations of existing system, survey of existing automatic railway gate controlling system is done. The system is redesigned by using a sensor near to the railway gate on track, that detects the arrival of the train and signals the closing of the gate. Similarly cameras are operational on crossing, generating the images of the track after gate closing to ensure all object removal from the track before crossing of the train during buffer time. Hence, this can be utilized in an unmanned railway gates which are more prone to accidents due to presence on track even after gate is closed.

Some facts studied to implement the concept are: Average speed [3] of Indian railway is 59km/hr to 91.82 km/hr. So base calculation reveals that ideal distance can be 3 km while signaling the train to stop or apply the brakes if some objects are still lagging behind on the track. If load sensor detects the arrival of train, it sends the signal to the microcontroller then the microcontroller activates the buzzer for warning the level crossing users that the railway gates are yet to be closed and the arrival of the train within a stipulated time. Camera sensor will observe the images and machine learning classifier, classifies the largest object and estimates the track clearing time. The controller then activates the pair of DC motors after the estimated time to control railway gates. After the train crossed level crossing, when second load sensor detects the departure of train, it sends the signal to controller. Microcontroller again activates the buzzer to notify that railway gates are yet to be opened. Controller activates in backward direction to open the gate.

B. Some existing example systems

There are various technologies available for automatic railway gate controlling system based on IR sensors, radar system, pressure sensors, ultrasonic sensors integrated on electronic board. The brief study of some good examples is stated as follows:

R.Ekalya[1,2018] suggested an efficient method to control the opening or closing of the railway gates using the help of Internet of Things(IoT) pressure sensors, embedded on Arudino board. Pressure sensors are bulky sensors and can be calibrated manually, so they are less accurate and prone to error. So the suggested system is less reliable and also obstacle detection between the level crossing gates at peak time is not handled.

Dr.S.Anila[7,2017] proposed a prototype for the automation of the level crossing gates at the railway station using IR sensors and intermediate obstacle detection using Ultrasonic sensors and GSM network. IR sensors work with line of sight between transmitter and receiver. The limitations of this system are: IR sensors are not able to distinguish between the objects with similar thermal energy

level irradiation. They are not cost effective. They can

control only one device at one time. They are not able to sense the objects which are not in LOS (Line of Sight) and above all they support shorter range and hence their performance degrade with longer distances. The sensing accuracy of the ultrasonic sensors is affected by climatic changes in the nature and hence they are not durable. These sensors also support short range of communication.

The paper proposes signal and data communication with GSM network which provide limited data rate capability and then complete system is dependent on GSM network availability. At time, this may delay the data transmission making entire system unreliable.

Pranav Sharma [8,2015] proposed a system using RFID, Pressure sensor, and servo motor to control mechanical movement of the gate. But it suffered from the drawback that it cannot be used for very high speed trains and also at the hilly areas where the pressure sensor may not perform accurately.

Greene R.J.[4,2006] stated about intelligent railway crossing control system for multiple tracks that features controller that receives message of trains by sensors and based on information on the message controller will close or open the railway crossing gates.

(3)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 9, Issue 10, October 2019)

But the feasibility approach is it is not possible to control the speed of train within one km so it can be extended to 3 km.

Jian-jua [3,2008] suggested sensor controlled the operation of gate according to sensors placed at distance. If any obstacle is detected red signal will be displayed for train indicating train to stop as it detects obstacle on track. Karthik krishnamurthi[5,2015] is sensor based automatic controlled laser beam and light dependent resistor to detect obstacle.

Theoretically, there exists a number of ways for the railway gate control system but every system has some drawbacks. Similarly only providing mechanical system may not work in the real world, it must be supported by image analysis and data analytics features. Automated railway gate at a level crossing is generally used to replace a gatekeeper who operates it manually. It manages two operations. First, it reduces the gate closing and waiting time. Second, it provides safety to the users and prevents accidents. The next section presents proposed system prepared by integrating some less complex but effective sensors and machine learning abilities.

III. PROPOSED SYSTEM

The major blocks in the proposed system consists of Microcontroller (model AT89C52), load sensor, camera sensors shown in the figures. Load sensor is a general purpose proximity sensor. The module consists of IR emitter and IR receiver pair. The module consists of LM358 IC. LM358 is a dual op-amp IC integrated with

two op-amps powered by a common power supply. It can be considered as one half of LM324 Quad op-amp which contains four op-amps with common power supply. The differential input voltage range can be equal to that of power supply voltage. The output of sensor is high and low otherwise. The status of sensor is checked by LED present on-board without additional hardware. The power consumption is low. It gives digital output. The proposed system uses buzzer for the alert signals.

The prototype of proposed system in given below. The system uses load sensor is used to detect obstacleand when train arrives gates are closed. The following system detects if any obstacle is present after closing of gates, it uses load sensor and measures the distance between it and the obstacle.

Whenever train touches base at the load sensor, transmitter transmits the signal to activate buzzer/light/alarm at the railway crossing so that the general population get s instruction that crossing will be shut and be away during buffer closure time.

At that point the control module initiates the motor and closes the gates on either side of the track. Once the train crosses, this module naturally lifts the gate. For mechanical operation of a gate DC adapted motors are utilized. As per the instructions produced at the microcontroller, the proper action of opening or closing the gate will be made. This logic was implemented in Embedded C and dumped to the Arduino Board. Also an indicator light and alarm to alert the vehicles/motorists/passersby about the approaching train is activated at crossing.

A. Projected module

The proposed system is consisting of three modules with primary elements as follows:

 Train module: consists of load sensor detecting train arrival and departure movement on the track available on Arduino board.

 Crossing control Software module: consists of machine learning to detect obstacle detection

 level crossing module works on crossing of train safely signaling the motors to control the movement of the gate.

IDE comprises of free software components, compiler, linker, debugger and serves as a single and unified GUI to microchip and hardware development tools

B. Elements

[image:3.612.364.556.552.663.2]

 Microcontroller AT89C52 Microcontroller is utilized in this structure. AT89C52 is a 8-bit microcontroller and has a place with Atmel's 8051 family AT89C52 has 8KB of Flash programmable and erasable read just memory (PEROM) and 256 bytes of RAM. AT89C52 has a perseverance of 1000 Write/Erase cycles which implies that it very well may be deleted and modified to a limit of multiple times.

(4)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 9, Issue 10, October 2019)

61

[image:4.612.88.254.214.339.2]

 Sensors The sensor faculties and reacts to occasions happening in the physical condition. It is equipped for estimating a physical marvel (like temperature, weight, warmth, moistness, etc) and changes it into an electric flag which can additionally be utilized for playing out a specific task.

Figure 2: Load Sensor

 In this system, load sensors are utilized for the equivalent sensing. A heap cell is a transducer that is utilized to make an electrical flag whose extent is legitimately corresponding to the heap being estimated. At least one burden cells can be utilized for detecting a solitary burden. More sensors are utilized for huge holders or stages, or high loads.

 Here crafted by the heap sensor is to detect the approaching burden and along these lines create electric signs. These signs are then sent to the microcontroller. Microcontroller forms these signs

 Stepper Motor A stepper engine, otherwise called venture engine or venturing engine, is a brushless DC electric engine that isolates a full revolution into various equivalent advances. The engine's position would then be able to be told to move and hold at one of these means with no position sensor for criticism (an open-circle controller), as long as the engine is painstakingly measured to the application in regard to torque and speed.

Figure 3: Stepper Motor

 Arduino Board Arduino board structures utilize an assortment of chip and controllers. The sheets are outfitted with sets of advanced and simple information/yield (I/O) sticks that might be interfaced to different extension sheets or breadboards (shields) and different circuits. The microcontrollers are normally modified utilizing a tongue of highlights from the programming dialects C and C++. Notwithstanding utilizing customary compiler toolchains, the Arduino venture gives an incorporated advancement condition (IDE) in view of the Processing language venture.

C. Process

While automating the railway gates operations, there is chance that vehicle may be locked between the crossing gates. Thus in order to save them, system has been taken some fruitful steps. Camera sensors have been placed at level crossing. It continuously transmits the images. Image could be programmed into controller. It is activated when gates are closed and so if any obstacle locked between railway gates and it is detected, it conveys the information.

Machine Learning algorithm implemented in the system near crossing to signal final gate closure to Arudino in case of obstacles on the track even after warning bell..

ab 

[image:4.612.327.574.422.699.2]

Figure 4: ARDUINO BOARD

(5)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 9, Issue 10, October 2019)

D. Obstacle Detections

 Object detection is the process of finding instances of real-world objects such as vehicles, bicycles, and persons, animals in images or videos. Object detection algorithms typically use extracted features and learning algorithms to recognize instances of an object category of a certain class.

 The first part is an algorithm for creating a feature vector (also known as a data point) given an image. A feature vector consists of several numbers that are measured or calculated from the image. These features are then used by the second part of the system, a machine learning algorithm, to classify the objects on the track to determine estimated time.

 K nearest neighbor the Euclidean distance with the weight of each feature normalized according to the variance of that features in the training data of the class vehicles, persons, animals.

 Cross-validation was used to estimate the percentage rates of correct and incorrect classifications that are needed for the analysis mentioned of images.

 Visual analysis and pattern recognition can be used to estimate the content of images. The example chosen for this project is classifying outdoor track images as vehicles, persons, animals.

 The high resolution cameras are available on railway gates. Periodically on interval of 10 msec-1sec, image of gap between two gates- track part between two gates is taken. The computer can be trained to detect and classify multiple objects within an image with high accuracy with the help of above mentioned machine learning.

 Object detection comprises of object localization refers to identifying the location of an object in the image. An object localization algorithm will output the coordinates of the location of an object with respect to the image.

 The popular algorithm from computer vision to localize an object in an image is to represent its location with the help of bounding boxes.

 This is one mechanism to get coordinates of objects in the image. Another approach for an object detection is to first build a classifier that can classify closely cropped images of an object. Neural network model is trained on a dataset of closely cropped images of a vehicle/or any object present on the track and the model predicts the probability of an image being a vehicle/pedestrian.

 To capture number of cropped images for more accurate detection, a sliding window mechanism may be applied. The size of the crop is the same as the size of the sliding window and repetitively image is cropped and passed to convolution network model or neural network model, which in turn predicts the probability of the cropped image is a vehicle/pedestrian.

The proposed system is highly reliable, effective and economical at dense traffic area, suburban area and the route where frequency of trains is more. Using automatic railway crossing system, we improve the rail road transportation facility by reducing the chances of occurrence of accidents at unmanned level crossings and providing immense safety. providing immense safety. Also this technique has fast operation than older system, it saves a lot of time as it is automated whereas manual systems take time for the line man to inform the station master to close and open the gate which will consume a considerable amount of time. The obstacle detection unit has been employed in the proposed system to reduce the accidents at level crossings upto maximum amount. Since the design is completely automated it can be used in remote areas where no station master or line man is present and it doesn’t degrade than existing system. Thus this system finds its applications in many cases.

IV. CONCLUSION

(6)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 9, Issue 10, October 2019)

63 REFERENCES

[1] R.Ekalya, Alex Pavithran KP, Manasi Biswas ,” Controlling Railway Gate Using Arduino”, Journal of Network Communications and Emerging Technologies (JNCET) Volume 8, Issue 4, April (2018). [2] R.Anuj Tyagi, Er.Goutam Arora “synopsis on automatic railway

crossing system” S.D. Institute of Technology & Management Israna-(Panipat). Evaluation of Structural and Biological systems V, SPIE (2006)

[3] Qiao Jian-hua, Li Lin-sheng, Zhang Jing-gang, “Design of Rai Surface Crack-detecting System Based on Linear CCD Sensor” IEEE Int. con on Networking, Sensing and Control, 2008.

[4] R.J. Greene, J.R. Yates, E.A. Patterson, "Rail Crack Detection: An Infrared Approach to In-service Track Monitoring", SEM Annual Conference & Exposition on Experimental and Applied Mechanics, 2006.

[5] Karthik Krishnamurthi, Monica Bobby, Vidya V, Edwin Baby, “Sensor based automatic control of railway gates”, International Journal of Advanced Research in Computer Engineering & Technology (IJARCET,) Volume 4, Issue 2, February 2015.

[6] .Hairong Dong, Bin Ning, Baigen Cai, and Zhongsheng Hou, “Automatic Train Control System Development and Simulation for High-Speed Railways”, IEEE Circuits and Systems Magazine, Volume 10, Issue 2, June 2010.

[7] Dr.S.Anila, B.Saranya, G.Kiruthikamani, P.Devi,” Intelligent System For Automatic Railway Gate Controlling And Obstacle Detection”, International Journal of Scientific Research & Growth • August 2017.

[8] Pranav Sharma,Rajesh Kumar,Sarika ,” Automatic Railway Gate Control System Based on RFID, pressure sensor and servo motor ”, Journal of Network Communications and Emerging Technologies (JNCET) ,Volume 5, Special Issue 2, December (2015).

Figure

Figure 1: Microcontroller
Figure 2: Load Sensor

References

Related documents

We retrospectively reviewed microbiological records for cerebrospinal fluid (CSF) and the medical records of patients with ABM admitted to the Chang Gung Memor- ial

It was decided that with the presence of such significant red flag signs that she should undergo advanced imaging, in this case an MRI, that revealed an underlying malignancy, which

If you reset your phone to its factory settings, all of your personal data stored on the phone, including information about your Google account, any other accounts that you have

Apply discount rate to lost earning capacity damages for each year to calculate annual present value of that year. – For past and current years discount factor is 1.0 – For

Extraordinary Recipients: MDS Activities of Daily Living Sum score of ten (10) or more and require special care or clinically complex care as recognized under the Medicare RUG

Oe. lutescens are constant types, but their characters are visibly inherited in the first generation through the gametes of one sex only. fastigiata and

The team effectiveness factors identified (team Goals and objectives, team leadership, team relationship, team roles and responsibilities, team communication, and trust and

The excess dipolar energy due to short range interaction between dipolar interactions between dissimilar molecules given by the value of 20.Swain 21