Design of Automatic Accident Response System and Traffic Signal Control at Intersection

Design of Automatic Accident Response System and Traffic Signal Control at Intersection Nakul Parashar, Preet Jain Electronics and Communication Dept, Shri Vaishnav Institute of Technology and Science, Indore, India nakul7391@gmail.com, preetjain@gmail.com Abstract: Fast response is very necessary when an accident takes place, due to increased traffic in cities it is very important that nearest hospital should be notified along with exact location of accident so that hospital can provide fast and efficient response. It is of utmost importance to reduce reaction time for accident, as well as time required to get the injured and bring them back to hospital. This paper presents a system which can find nearest hospital from the location of accident; notify them exact location of accident site, display shortest route on web browser and turn the red traffic signal light green when ambulance is nearing the intersection. The designed system requires vehicles to be installed with a unit which can provide coordinates at the time of accident. Keywords: Google Maps API, GPS, GSM/GPRS, Wi-Fi, Server, Accidents, Ambulance. I. INTRODUCTION The population in cities is increasing and so are the vehicles. With growing vehicles the accidents are also increasing. The ministry of road transport and highways reports [1] 4, 86,476 accidents in 2013 in which 1, 37,572 people have died. 16 people die every 3.7 minutes which means 390 deaths every day. In order to reduce this toll fast response should be provided to injure. The response time becomes very large when hospitals has no or incorrect information regarding the location of accident or the information misinterpreted or passed to hospital which is far from accident. Highway authorities have to rely on information provided by witnesses and passersby which is often incomplete. To worsen the situation traffic signals with cluttered vehicles, damaged roads, small roads increase the time required by ambulance to reach accident location and head back to hospital. Consider a situation where an ambulance is waiting in long queue for green signal and there is not one signal but there could be many signals in the path which can force the ambulance to stop and thereby increase the total time required to provide assistance to victims and take them back to hospital. This paper introduces a server based system AARS (Automatic Accident Response System) which can noticeably reduce reaction time and response in road accident scenario. The AARS can automatically detect accidents, provided that vehicles should have a unit installed which can provide the location in the form of latitude and longitudes to server at the time of accident, or a Smartphone can also be used to do this. The AARS is capable to find the nearest hospital and supply that hospital the location of the accident via email or SMS to reduce notification time. It can also provide shortest route to accident location using Google Map API for ambulance operator. With the use of GPS receiver and GPRS it can monitor ambulance of interest. AARS calculates the distance of ambulance from traffic signals in its path and if the ambulance is nearing the intersection traffic signal at that intersection can be turned green to provide seamless path to ambulance to and fro. This can be modified to create green corridor automatically. Traffic signal are connected to server using Wi-Fi and work on interrupt provided by server. Automatic system aims to save human lives and reduce human efforts. The paper is structured as follows. Section II presents similar research in this area. Section III describes the system architecture. Section IV describes AARS Section V gives hardware description of ambulance unit and traffic signal. In section VI paper is concluded and future refinements are discussed in section VII. II. LITERATURE SURVEY The traffic system and emergency response is the field of ongoing research. Similar projects which deals with minimizing reaction time at emergency situations like accidents and as a replacement to conventional approach to notify hospital and also to reduce time required by emergency vehicles are discussed here. Various sensors and algorithms are used to provide real time data about the traffic. To deal with the problem of cluttered vehicles at intersection [3] a traffic monitoring system is developed. This system uses cameras at intersection which by application of markov random field algorithm can determine occlusion situation of vehicles, this method mainly focus on intersection because at intersection it is difficult to track cluttered vehicles. The system basically monitors traffic flow at intersection. A software platform to handle all type of emergencies is designed [4] which has some similarities to e-notify system [6], system automatically generates plan and provide a mean to display the scene using GIS to traffic operator who also has access to resource management and can manages resources. Scene information system can efficiently display emergency situation. The software 10

ACCIDENT International Journal of Electrical Electronics & Computer Science Engineering platform is very efficient to reduce response time in emergency situation. In emergency situations to reduce time it is necessary that emergency vehicle will try to avoid high traffic roads and safely navigate through intersection. Taking these two points into consideration a routing system is developed [5]. A data collection unit using RTCU-Mx2i [10] is used for emergency vehicles and traffic lights. A routing algorithm is developed which uses neural network to update real time traffic data along with the control of traffic light. To reduce significant response time e-notify system is implemented using vehicular networks (V2V) or VANET (Vehicular Ad hoc NETworks) [6] this system is able to detect accidents. In the proposed system every vehicle should have On Board Unit (OBU), which can detect the impact by collecting information from sensors of vehicle and sends warning message to control centre. This system can effectively reduce the response time of emergency services by fast detection of traffic accidents. This system is similar to our designed system which also includes the capability of traffic light control similar to learning routing system to further reduce time. Traffic light plays a role to provide fast response to emergency situations, traffic lights use timers and are not meant to handle emergency situation. In Kuwait flexible and reprogrammable traffic light system is implemented [7]. The system is referred to as Emergency Traffic Light (ETL). The ETL use RF communication which is divided into two parts sender and receiver, the sender unit is manually operated by ambulance driver. III. SYSTEM ARCHITECTURE Figure 1 shows the general architecture of the designed system. Server is the main controlling unit which detects the happening of accident. It is connected to ambulance unit which is a GPS receiver connected to GPRS modem used to send real time location of ambulance, it is also connected to traffic signals of the city using Wi-Fi and interrupt a particular traffic signal which is in vicinity to ambulance and make it green. The purpose of the system is threefold. 1. Sort through database to find nearest hospital from accident site and notify that hospital the exact location of accident. 2. Show graphically shortest route to accident location which can be accessed by anyone. 3. Track the ambulance to accident site and then to hospital and clear red traffic signal in its path. Figure1. System Architecture The vehicle should have the means to communicate with the server using GSM or any other way other than that Smartphone can also be used to communicate with server to provide coordinates of accident location. Server (AARS) in this system is the controller which is connected to traffic signal and connected to various real time databases. AARS can run automatically on any Linux or window based system and even ambulance operator can see shortest route as well as real time location on their Smartphone by entering relevant URL. System also uses an ambulance unit which is used to update its location frequently on the server to track the ambulance. Traffic controller is connected to AARS via Wi-Fi enabled device and AARS has full control over traffic signal. IV. AARS AARS gets the information about accident location in the form of latitude and longitude. It is assumed that vehicle has the means to update the database at the time of accident, any type of IoT device or Smartphone with GPS receiver can be used. Figure 2 shows the functional block diagram of the AARS. ACCIDENT LOCATION AARS NEAREST HOSPITA Figure 2. Functional Diagram of Server WEB BROWSE R Traffic Signal 11

It is the main controlling unit hence handles all the events. It frequently checks the database for new location. When it gets a location it searches the entire database of hospitals to find the nearest hospital and display the shortest possible route along with informing the hospital exact location of accident site. The calculation and visualization is done using Google Maps API [2] works on JavaScript to embed in web page. The shortest route can be viewed by ambulance operator. An email or SMS is used to supply necessary information to hospital. From the beginning our main focus is on reducing time, the process of finding nearest route and forwarding message to hospital takes approximately 5 seconds. Once the hospital gets information they can dispatch ambulance to accident site. Figure 3. When Accident Detected Figure 3 shows the output form AARS after the accident is detected, once the accident detected it creates another instance automatically and waits for accidental information to occur. It acts as background process after accident detected and checks for location at fixed interval. Figure 4 shows the response of the server after notifying nearest hospital, it is the shortest path to accident site including address of location and time to reach, opens after 5 seconds which is sufficient for Google servers to respond with shortest distance, route and address of accident location and hospital. It is like a foreground process launched by server to track the ambulance and control traffic lights. Tracking ambulance up to hospital is a lengthy process and system should be able to respond to multiple accidents hence another instance is created as a foreground process. The output can be viewed on web browser. So a background page which is the page which detects accidents can launch many foreground instances to track many ambulances and control various traffic lights. Markers are used to show accident location, nearest hospital and traffic signal. It can be used by ambulance operator to navigate to accident site. The accident location can also fed manually in order to make green corridor. A unit on ambulance is used to provide frequent location update to database and AARS uses this database to display and track position of ambulance. Figure 5 shows the real time tracking of ambulance by a red marker on the map and it gets updated according to ambulance s movement. AJAX is used here to stop automatic reloading of page with each new location of ambulance. The final job for the AARS is to preempt obstruction in the path of ambulance at intersections. Figure 4. Response of Server After Finding Nearest Hospital 12 Figure 5. Real Time Tracking of Ambulance A red signal means ambulance has to stop and wait in long queue on accumulated vehicles. AARS takes control over the traffic signal which is in the vicinity to ambulance. AARS calculates distance between all the traffic signals and sends an interrupt signal with the direction of the path to be cleared to that signal which has the distance less than 200 meters or more from ambulance to provide

From server To server International Journal of Electrical Electronics & Computer Science Engineering sufficient time for long queue of cluttered vehicles to clear the intersection. Figure 6 shows the response when ambulance reaches to traffic signal, at this instant server sends interrupt signal to traffic signal along with which signal to be cleared NS or EW. Fig.6 Ambulance Nearing Traffic Signal V. HARDWARE SYSTEM Along with AARS hardware units are required at each traffic signal and on board each ambulance. Unit on board the ambulance is a simple GPS receiver interfaced with GPRS modem using a controller to transmit its location to server frequently. Traffic signals are preprogrammed using controller and connected to AARS using Wi-Fi. A. Ambulance Unit: The system requires a unit on all the ambulance to send their location to server. Hospital s work is finished after dispatching ambulance, once the ambulance start heading towards accident site it needs to reach as fast as possible to the site. It is highly improbable that roads are empty also every traffic signal ambulance encounter would be green and it is highly likely that it has to wait for signal to be green. Main focus of this project is to reduce time to save victims, the reaction time that is time to inform hospital is reduced to maximum of 5 seconds, reduction in time required for operation of ambulance is only possible by controlling traffic lights in the vicinity of ambulance and for that it is necessary to know real time location of ambulance which is possible only when ambulance has some means to send its location, since Wi-Fi is not available everywhere in the present working region GSM is the only possible mean to communicate with server. Figure 7 shows the block diagram of ambulance unit. A unit on board ambulance uses a GSM/GPRS modem. Sim900 is used for this purpose. A GPS receiver L80 is used to receive latitude and longitude of ambulance. Both the GPS and GSM are controlled by a microcontroller unit which first gets latitude and longitudes from GPS and then fed them to GSM which transmits them to server where they can be used to calculate distance from signals. B. Traffic Signal: Traffic signals have fixed location and it is possible to have Wi-Fi connectivity around it, GSM could be used but the cheaper alternative is using any Wi-Fi enabled IoT device. Traffic signal contains Wi-Fi enabled ESP8266-12F [9] it has Tensilica xtensa LX106 core which is used for IoT applications. Traffic controller uses 8051 family based microcontroller AT89C2051.Microcontroller is used to perform foreground operation of traffic control. Green, red and yellow LEDs are connected to ports for controlling North-south and East-west traffic. Figure 8 shows block diagram of traffic controller. Traffic Lights Sim900A GSM/GPRS modem Figure 7. Ambulance Unit Microcontroller unit L80 GPS receiver ESP8266-12F Wi-Fi Module AT89C2051 for Traffic Control Figure 8. Block Diagram of Traffic Control Normal traffic control works in foreground while Wi-Fi module is used in background to interrupt the microcontroller. Traffic signal when turn on provide its IP address and coordinates to server, then start executing preprogrammed instructions and Wi-Fi module is set in listener mode. AARS uses coordinates to calculate its 13

distance with ambulance s real time location. When server needs to interrupt the traffic signal which is nearer to ambulance it sends a signal to Wi-Fi device along with the direction NS or EW which is to be cleared and keep the green light on for a predefined time to completely clear the intersection. Wi-Fi device is based on NodeMCU firmware which is an open source platform that uses LUA scripting language. VI. 14 CONCLUSION AND RESULTS The main objective of this project is to minimize time required by emergency vehicles to respond to accidents. Right now hospitals gets information by witnesses and passersby and ambulance can be called by helpline number [8]. Comparing with the first scenario reaction time is large which is reduced to 5 seconds and can be reduced more since it is fully based on internet speed and connectivity. Now in second case ambulance has to wait at traffic signal unless any traffic police personnel are there to clear the traffic which is a time consuming process. Using the AARS traffic lights are controlled automatically which minimize the time spent at intersection. Like the e-notify system [6] this system can automatically gets accident location provided that vehicle should have an On Board Unit installed also V2V or V2I networks can be used to provide location information to database, traffic lights when connected to local area network can reduce response time significantly so there is no need for RF communication [7] GPS unit in ambulance can automatically provide its location time to time for real time tracking so driver doesn t have to do anything. The driver can also view shortest route on web browser in a Smartphone. Google Maps [2] provide almost accurate results and maps are updated frequently so there is no need for any major change in algorithm, also no operator support means it can run 24*7 without any operational and maintenance cost. It can also control traffic signals without any automatic accident detection to make green corridor hence it is a cheaper and effortless alternative to green corridor by manually feeding location to airport or another hospital. Traffic signal unit ESP8266 [9] is cheaper and can be used at large scale. VII. FUTURE WORK Web based system is an efficient way to provide fast response and automation. This system can be modified to keep track of locations of ambulances and can be used to directly inform ambulance instead of hospital to further reduce time as a replacement of 108 emergency services. Smartphone application can be built to show shortest route automatically removing the process of manually feeding URL to web browser. Along with accidents more emergency services can be added like fire or police assistance. VIII. REFERENCES [1] Road Accidents in India 2013 Government Of India, Ministry Of Road Transport & Highways Transport Research Wing New Delhi http://www.morth.nic.in [2] Google Maps JavaScript API available at https://developers.google.com/maps/documentation/j avascript/ [3] Kamijo, S.; Matsushita, Y.; Ikeuchi, K.; Sakauchi, M., "Traffic monitoring and accident detection at intersections," in Intelligent Transportation Systems, IEEE Transactions on, vol.1, no.2, pp.108-118, Jun 2000 [4] Guolin Wang; Deyun Xiao; Gu, J., "Design and implementation on software platform of emergency management system for traffic incidents," in Electrical and Computer Engineering, 2008. CCECE 2008. Canadian Conference on, vol., no., pp.000475-000478, 4-7 May 2008 [5] Vlad, R.C.; Morel, C.; Morel, J.Y.; Vlad, S., "A learning real-time routing system for emergency vehicles," in Automation, Quality and Testing, Robotics, 2008. AQTR 2008. IEEE International Conference on, vol.3, no., pp.390-395, 22-25 May 2008 [6] Fogue, M.; Garrido, P.; Martinez, F.J.; Cano, J.-C.; Calafate, C.T.; Manzoni, P.; Sanchez, M., "Prototyping an automatic notification scheme for traffic accidents in vehicular networks," in Wireless Days (WD), 2011 IFIP, vol., no., pp.1-5, 10-12 Oct. 2011 [7] Nasser Al-Ostath, Zainab Al-Roudhan, Fatma Selityn, Mohammed El-abd, implementation of an emergency vehicle to traffic lights communication system [8] http://www.nhp.gov.in/helpline_pg [9] http://www.esp8266.com/ [10] Logic IO, Documentation of the RTCU unit, available at http://www.rtcu.dk/rtcu_products.htm#mx2pro