- freely available
Electronics 2019, 8(8), 896; https://doi.org/10.3390/electronics8080896
- Providing medical help to victims of road accidents in a timely manner.
- Notification about the precise situation to the first responders.
- Lack of a permanent database, holding all appropriate documents and records, which can be examined as and when necessary.
- On-Board Sensors: It is capable of accurate road accident detection through an OBU, consisting of an accelerometer and a gyroscope along with identification of an accident event through Global Positioning System (GPS). It has a provision to capture images using a high-definition camera module from inside a car and then attach the captured images/data along with the other information that is sent to the destination. The complete message is routed to the edge node using a vehicular ad hoc network.
- Edge Gateway: These are the entities in all the regions that are positioned in the control room. It has the responsibility to receive an accident notification and then process the sensed data. Another responsibility of Edge Gateway is to distinguish the number of casualties and the person involved in the accidents using face detection procedures. Furthermore, it processes the information received from the nodes in the network and then informs the hospital so that it can immediately dispatch the ambulance. Moreover, it has the ability to store information temporarily prior to receiving securely by the cloud server.
- Cloud Platform: The last but not the least tier of the system has a responsibility to store and process information received from the edge nodes. Here, a visualization tool is used that generates a query to fetch the information from the database. It is further used to study and analyze road accident data for various functionalities such as for policy-making, including research and documentation.
- The first procedure deals with the automatic accident detection mechanism that also provides a module to record the necessary information. Additionally, the system based on the implemented TestBed deals with the following three novel aspects:
- The detection of an accident with the help of sensed information through appropriate sensors where an algorithm has been developed to gather and monitor the required information. Based on the sensed information, it performs the logical computation to detect the accident. Once it is detected, it generates an alert message. This article includes the details of data processing of the accident detection module in a practical manner. Additionally, the contribution is presented through a flow chart of the proposed system to detect an accident correctly and in a timely manner.
- The second module is used to identify the location of the accident and medical service nearby. It then notifies the hospital not only about the condition of the person but also about the severity of the event based on the damage. Moreover, it stores useful information about the incident for future use.
- The third part deals with the hospital module, which not only sends the ambulance to the incident location on emergency alert, but also gathers the passenger’s status like injuries, deaths along with medical proof of alcohol and personal details like name, age, and the number of passengers, etc. This information is then sent back to another module for future use.
- The second procedure deals with the analysis of stored information like the location of accidents, reasons for accidents, injuries and death statistics of the accidents that occurred in the specific region to find the reasons for accidents. The system is tested in a lab environment, and, for this, we used a library that generates data on ‘a’ and ‘c’ module mentioned in procedure-1. Then, that data are transferred using the implemented system (TestBed) and analyzed using visualization techniques. Moreover, we took the recorded realistic data of a city and we generated graphs using our proposed system to analyze the reasoning. As far as the visualization part is concerned, it enables the authorities to make proper decisions to tackle the incident and to take admissible precautions. Additionally, for the visualization part, raw data have been converted into information by passing it through the proposed system that has been developed in the Lab as a TestBed. Therefore, one of the contributions of this work is that it has created a meaningful visual output using raw data at the Cloud.
2. Literature Review
3. Proposed System
- instant detection of accidents automatically (without the involvement of any human)
- precise identification of the position of an accident,
- endowment of an exact and well-timed medical facility.
- Tier 1: On-board Sensors This tier of the system consists of on-board sensors as shown in Figure 4. In order to detect accidents, the in-vehicle OBUs constantly monitor the acceleration and orientation of the vehicle. When an accident happens, the most essential data, i.e., the date and time, the geographical location (latitude and longitude), the chassis number, and the inside temperature of the vehicle, are sent to the nearby Edge node.Tier 1 is responsible for handling two main phases: firstly, automatic accident detection and emergency alert and, secondly, accident management as presented in Figure 5. The proposed system architecture is deployed in each particular zone as depicted in Figure 3, where it is handled independently in each zone. However, data are recorded in the control room of each zone, which is transferred via an Edge node deployed in each zone and further handled at the cloud. Each vehicle in a particular zone of the system is equipped with an OBU, where Raspberry-Pi, GPS, IMU, and camera modules are embedded on it. These modules help to detect the accident, and the processing unit of OBU is used to process the gathered information. When an accident is detected, it generates an alert message to the deployed control room in a particular region using Vehicle-to Vehicle (V2V) communication. The control room records the accidental data and sends an emergency alert to the hospital for the first-aid service (ambulance).An incident like a road accident could either be a collision with other vehicles or any other object or a rollover. Whenever a collision happens, the speed of the vehicle decreases dramatically, and its acceleration can be measured by the accelerometer. Roll-overs can be detected by monitoring the angular position of the vehicle or by using angular velocities. The angular position of the vehicle is computed by using the data gathered from a gyroscope. In case of a roll-over, an emergency message is generated without incorporating the data from a glass break detector and pulse sensor, due to its severity; however, in case of collision, data from all the sensors are analyzed. The flowchart of the accident module is shown in Figure 6. It can be seen that all the sensors are initialized and updated at regular intervals. If the collision detection is above the predefined threshold, the emergency alert is sent to the rescue services and to the nearby vehicles. Contrary to this, if the collision detection by the IMU does not exceed the threshold value, then the system will fetch data from the sound sensor. If the sensor circuitry detects a noise greater than the threshold value, then the emergency alert is sent. If there is no noise heard by the sensor circuitry, then the system will take the data from the pulse sensor. If the heart rate is above a certain threshold, then the alert message will be sent; otherwise, an alarm is generated that is usually canceled by the driver in normal circumstances. If the driver does not turn off the alarm, it is assumed that the driver is not in a condition to do so, and again an emergency alert is sent. Location coordinates and image of the vehicle from inside are fetched from the GPS module and camera module, respectively, and then it is enclosed in the emergency message.When an accident is reported for an emergency service to a hospital, it dispatches an ambulance to the reported location quickly. Because of the incomplete information of the exact situation of the victims and lack of information regarding the number of injured persons, the paramedic staff may find it difficult to make any judgment regarding the first aid services. Instead, the additional information of a specific calamity, a better emergency medical service along with advanced technology support, first aid kits, tools, and medicines can be sent to the victims. Additionally, it would assist with providing timely help to the victims. To cope with it, a high-definition (HD) camera is integrated into the proposed system to capture the instant when the accident occurred. A Logitech HD Universal Serial Bus (USB) C270 Webcam that is capable of capturing 720 pixel, 30 (frames per second (fps)) wide-screen images  was used.
- Tier 2: IoT Gateway All of the smart devices are mounted on-board deployed inside the vehicle; therefore, it is more hands-on to handle and process the recorded data at the Edge. In this case, the Edge node is closer to the source of data-generation, rather than at the distant cloud. Additionally, it is not needed to send all the recorded data to the cloud, but just the refined data can be sent, thereby saving network bandwidth. Data packets traversed via relay vehicles in a multi-hop ad hoc network, connecting a Wi-Fi module of an OBU inside the vehicle to other vehicles in the neighborhood. Additionally, this communication is facilitated by an IEEE 802.11n Wi-Fi USB dongle in ad hoc mode. After traversing the whole path leading to the Edge node, the sent data shall be received at the gateway. The IoT node will then perform various tasks assigned to it, i.e., the segregation of received data to extract the useful information, storage of the associated picture, and forward transmission of the processed data. These are further conferred as follows:
- Facial Detection: To identify the number of victims inside a vehicle when an accident happens, the facial-detection is acquired from the image encapsulated in the received message using Open Source Computer Vision Library (OpenCV). The state of victims like their position or stature at the time that the misfortune happened would provide helpful actualities, which would be useful for the medical team who will come to the incident. By having such actualities, the first responder can forecast how long the accident victims can endure without any medical assistance .
- Data Preprocessing: Sensors deployed on the OBU monitor the required parameters and acquire data when it is needed. To avoid the delay involved, the acquired data are sent to the Edge node. It is highly desirable that the recorded data from the accident location and hospital are processed closer to the edge of the network i.e., control room, before transmitting it to the cloud for long-term storage. There are two advantages; firstly, it will clean the data of any incongruities or bad information beforehand to save network bandwidth, and, secondly, the data will be wrapped in a way that is easier for the end-user at the cloud to interpret. Additionally, it avoids the extra-processing on the recorded data at the cloud. Moreover, provisional storage can be provided for the data at this node before transferring it to the cloud to avoid loss of data due to the dynamic nature of the network between the sensor nodes deployed in the mobile node and the Cloud.Once the required information is processed close to the Edge node, it is securely cached. Then, the processed information is transferred to the central server at the cloud. In addition, the communication of edge to the cloud platform is based upon Transmission Control Protocol (TCP) because of its connection-oriented nature and reliability. This communication of the edge node from the control room to the central server at the cloud is done using the Internet as the cloud service is universally accessible through the Internet.
- Tier 3: Cloud Platform The Central Control Unit (CCU) accepts accident alert notifications and takes appropriate actions. The CCU is responsible for data retrieval from the received data packets. Additionally, it provides data storage and handles visualization with the help of a front-end tool, also installed on the Cloud.To test the implemented system on the cloud, a testbed is needed. A Testbed requires an Infrastructure as a service (IaaS) model, but, to the best of our knowledge, none of the former stated commercial service providers seemed to select it as a practical option. Consequently, a private cloud is used to implement and test the designed proposed system. The commercial cloud computing services are mostly used by start-up businesses or medium-sized enterprises to cut the costs of building their own applications, hosting on third party web-servers and maintenance. Therefore, a web server application called Apache is used to build a personal web server on a Linux-based operating system. Apache can serve HTML (HyperText Markup Language) files over HTTP (HyperText Transfer Protocol) on its own and can serve dynamic web pages using scripting languages such as PHP (Hypertext Preprocessor), with the help of additional modules. The data collected from the sensors in the Testbed may provide insightful information about the occurrences of accidents over time, in different parts of the region. Over time, when an accident occurs, its location may be identified, and effective preventive measures may be worked out, to at least lessen their occurrences, if not to completely get rid of them. In case strong preventive measures appear to be hard to implement, then at least the emergency services like ambulances and paramedic staff can be stationed nearby by identifying high-risk accident locations for the rescue operation.Simple and rather structured data representing concrete values of accident events are acquired by the sensors. To make a comprehensive and easily-understandable analysis of this data, an open-source Relational Database Management System (RDBMS) was used with PostgreSQL for data management. PHP is used for server-side scripting so that the relevant personnel administrating the CCU is able to customize the response according to their needs, and conveniently interact with the database.
- The accident detection module detects an accident event with the help of IMU MPU6050 and the OBU takes geographical information (latitude and longitude) from the installed GPS sensor of OBU. This way, the system serves as a “black box” that can help to indicate the location of an accident, cf., in Section 3.
- When an event of an accident is detected successfully, the camera module of the system takes an image of an inner view of the vehicle involved in the crash. The acquired data information is transferred to the control room where an Edge node is deployed to handle particular data as discussed in Section 3 Tier 1. It helps in providing the inner view of the vehicle for crucial situational awareness of the victims.
- The pre-processing of the data, like identifying the number of victims via facial detection, useful data extraction and sorting, is handled at the Edge node. It also handles the temporary storage of data. Meanwhile, the server in the control room sends an emergency alert to a hospital close to the accident as discussed in Section 3 Tier 2.
- Upon reception of an alert, the hospital dispatches an ambulance along with the medical team to the location of an accident. The paramedical staff is responsible for providing first-aid to the victims at the accident location and bringing them to the medical centers, if necessary.
- The information of the accident location, details of the vehicles including drivers and the status of victims are important for the traffic department. Therefore, the traffic department carries out its standard protocol and acquired useful data about the accident and the driver credentials.
- Finally, the aggregated data information is recorded into the central database at the cloud platform. The recorded data can be retrieved for the desired accident information whenever it is needed. The implementation of this process is done at Tier 3 of the proposed system architecture, cf., Section 3 Tier 3.
4. System Validation and Results
4.1. Accident Detection Phase
4.2. Communication Phase
4.3. Performance Analysis of the System
4.4. Data Analysis of Road Accidents: Real World Statistics
4.5. Accidents Pattern Analysis on Different Types of Roads
- Analysis on Federal Roads:Figure 18 and Figure 19 present the fact that varying levels of injuries caused inside the built-up areas on federal roads are higher than those caused outside of it. One main reason can be the presence of a number of activities inside the residential or commercial places when compared with the outskirts. Therefore, more accidents occur inside these populated places rather than outside of them. Though less in number, the accidents that do occur outside the city limits tend to be more brutal and severe, thus causing greater damage.The recorded data on federal roads inside the build-up areas of era 1991–2017 presented in Figure 18 shows that there were 1,044,451 cases of personal injuries, among them, 1,149,046 people faced slight injuries. Similarly, the recorded data on federal roads outside the built-up areas in the same era presented in Figure 19 show that there were 858,154 cases of injuries, among them, 936,236 faced slight injuries. The difference between the numbers reverse, when the persons killed or severely wounded as a result of these accidents, are counted. There were 10,408 cases of death in the first scenario while 41,869 in other cases, which is a four-fold increase. Similarly, there were 208,959 people who faced serious injuries inside the built-up area while 357,817 were in other cases. Hence, there might be more accidents in cities or towns, but, outside of cities, these accidents are considered more dangerous due to the lack of facilities. At the same time, it is good to notice that German authorities are carrying out their duties rather effectively, and there is a stable decrease in the number of casualties and injuries over the years.
- Analysis on District Roads:Figure 20 and Figure 21 present a similar scenario of road accidents’ recorded data for the years 1991–2017 on the district roads. Although the total number of casualties and injuries on district roads is much less when compared to federal roads, the drop in personal wounds or slight injuries over the last decade is not very satisfying, especially in the case of roads inside the built-up areas. For instance, on the district roads inside the built-up areas, the number of personal injuries was 16,526 in the year 2008 while 15,997 in the year 2017, which is a mere decrease of 529 cases over a decade. Similarly, the number of people who faced mild injuries was 17,788 in 2008 and 17,109 in 2017, which is again a decrease of 679 cases. Although the overall trend is positive, a sharp decrease in these figures referenced is needed, especially at this period of time, when the roads in Germany are considered to be the safest.The number of people who died as a result of a road accident inside built-up areas on a federal road in the year 1991 is 943, and for the year 2017 is 134, which is a seven-fold decrease. Similarly, the persons who died outside the built-up areas on a federal road in 1991 is 2713, compared to 688 for 2017, which is almost a drop of four times. The same is true for the district roads. Moreover, 325 and 999 people died in road accidents in the year 1991 inside and outside the built-up areas on district roads, respectively. The similar statistics are observed in 2017, hence indicating a decrease of around one-third.
- Analysis on Provisional Roads:Figure 22 and Figure 23 illustrate the number of casualties and injuries that occurred inside and outside the built-up areas on provincial roads, which indicates a greater loss than occurred on district roads, but less than the federal roads. A positive trend of dropping the number of injuries is observed inside and outside of the built-up areas 2015 and 2017. Although the trend in Figure 22 and Figure 23 might give a different impression, the total number of personal injuries on both types of these roads is very similar.
- Analysis on the Federal Motorway: The federal motorway (Bundesautobahn) is the federally-controlled-access highway system throughout all of Germany. For some classes of vehicles, this motorway has no speed limit, which makes it unique and popular throughout the world. Figure 24 presents the recorded data of federal motorways; it is observed that the number of personal injuries was decreasing up until 2012; then, it began to grow; since then, the number of personal injuries was increased in 2017. Similarly, the people who were slightly and severely injured or died in a road accident have also increased in 2017.The reason might be the increasing economic activity, as Germany is now the world’s fourth largest and Europe’s biggest economy. The most visual proof of its financial activity is the number of large logistics vehicles that carry goods and supplies to-and-fro, on its highways and motorways. In order to postulate a more convincing theory, further analysis of this subject is required.
4.6. Major Causes of Road Accidents Based on Types of Vehicles
- All occupants inside a car must wear seat belts, including those in the back seat.
- All children who are 150 cm or less in height must be placed in a child seat.
- No phone calls or chat application should be taken/used while driving.
- All speed limits on different parts of the road must be respected.
- Driving is not allowed while under the influence of alcohol.
Conflicts of Interest
|API||Application Program Interface|
|CCU||Central Control Unit|
|E-HAMC||Emergency Help Alert Mobile Cloud|
|GPS||Global Positioning System|
|HTML||Hyper Text Markup Language|
|HTTP||Hyper Text Transfer Protocol|
|IoT||Internet of Things|
|IoV||Internet of vehicles|
|IaaS||Infrastructure as a Service|
|OpenCV||Open Source Computer Vision Library|
|WHO||World Health Organization|
|RDBMS||Relational Database Management System|
|TCP||Transmission Control Protocol|
|USB||Universal Serial Bus|
|VANETs||Vehicular Ad hoc Networks|
- Save Lives—A Road Safety Technical Package. Available online: http://apps.who.int/iris/bitstream/handle/10665/255199/9789241511704-eng.pdf (accessed on 14 January 2019).
- Road Traffic Injuries. Available online: http://www.who.int/en/news-room/fact-sheets/detail/road-traffic-injuries (accessed on 10 January 2019).
- Khaliq, K.A.; Qayyum, A.; Pannek, J. Prototype of Automatic Accident Detection and Management in Vehicular Environment Using VANET and IoT. In Proceedings of the 2017 11th International Conference on Software, Knowledge, Information Management and Applications (SKIMA), Malabe, Sri Lanka, 6–8 December 2017; pp. 1–7. [Google Scholar]
- Khaliq, K.A.; Raza, S.M.; Chughtai, O.; Qayyum, A.; Pannek, J. Experimental validation of an accident detection and management application in vehicular environment. Comput. Electr. Eng. 2018, 71, 137–150. [Google Scholar] [CrossRef]
- Ray, P.P. A Survey on Internet of Things Architectures. J. King Saud Univ.-Comput. Inf. Sci. 2018, 30, 291–319. [Google Scholar]
- Da Silva, C.; Rodrigo, A.; da Fonseca, S.; Nelson, L. On the Location of Fog Nodes in Fog-Cloud Infrastructures. Sensors 2019, 19, 2445. [Google Scholar] [CrossRef] [PubMed]
- Morabito, R.; Petrolo, R.; Loscri, V.; Mitton, N. LEGIoT: A Lightweight Edge Gateway for the Internet of Things. Future Gener. Comput. Syst. 2018, 81, 1–15. [Google Scholar] [CrossRef]
- Lucovsky, M.; Collison, D.; Spivak, V.; LADDAD, R. Cloud Platform Architecture. US Patent 8,627,426, 7 January 2014. [Google Scholar]
- Ahmed, E.; Ahmed, A.; Yaqoob, I.; Shuja, J.; Gani, A.; Imran, M.; Shoaib, M. Bringing Computation Closer Toward the User Network: Is Edge Computing the Solution? IEEE Commun. Mag. 2017, 55, 138–144. [Google Scholar] [CrossRef]
- Raut, P.; Sachdev, V. Car Accident Notification System based on Internet of Things. Int. J. Comput. Appl. 2014, 107, 29–31. [Google Scholar] [CrossRef]
- Chen, L.; Englund, C. Every Second Counts: Integrating Edge Computing and Service Oriented Architecture for Automatic Emergency Management. J. Adv. Transp. 2018, 2018, 7592926. [Google Scholar] [CrossRef]
- Fogue, M.; Garrido, P.; Martinez, F.J.; Cano, J.C.; Calafate, C.T.; Manzoni, P. Automatic Accident Detection: Assistance through Communication Technologies and Vehicles. IEEE Veh. Technol. Mag. 2012, 7, 90–100. [Google Scholar] [CrossRef]
- Khaliq, K.A.; Qayyum, A.; Pannek, J. Synergies of Advanced Technologies and Role of VANET in Logistics and Transportation. Int. J. Adv. Comput. Sci. Appl. 2016, 7, 359–369. [Google Scholar]
- Yang, F.; Wang, S.; Li, J.; Liu, Z.; Sun, Q. An Overview of Internet of Vehicles. China Commun. 2014, 11, 1–15. [Google Scholar] [CrossRef]
- Darwish, T.S.; Bakar, K.A. Fog Based Intelligent Transportation Big Data Analytics in the Internet of Vehicles Environment: Motivations, Architecture, Challenges, and Critical Issues. IEEE Access 2018, 6, 15679–15701. [Google Scholar] [CrossRef]
- Al Najada, H.; Mahgoub, I. Anticipation and alert system of congestion and accidents in VANET using Big Data analysis for Intelligent Transportation Systems. In Proceedings of the IEEE Symposium Series on Computational Intelligence (SSCI), Athens, Greece, 6–9 December 2016; pp. 1–8. [Google Scholar]
- Aazam, M.; Huh, E.N. E-HAMC: Leveraging Fog Computing for Emergency Alert Service. In Proceedings of the IEEE International Conference on Pervasive Computing and Communication Workshops (PerCom Workshops), St. Louis, MO, USA, 23–27 March 2015; pp. 518–523. [Google Scholar]
- Logitech C270 HD Webcam. Available online: https://www.logitech.com/assets/46735/2/hd-webcam-c270.pdf (accessed on 23 December 2018).
- White, J.; Thompson, C.; Turner, H.; Dougherty, B.; Schmidt, D.C. Wreckwatch: Automatic Traffic Accident Detection and Notification with Smartphones. Mob. Netw. Appl. 2011, 16, 285–303. [Google Scholar] [CrossRef]
- Khaliq, K.A. Emergency Disaster Management. Available online: https://www.researchgate.net/publication/335146412_Configuration_of_Ad_Hoc_Networks_on_Raspberry_Pi_for_Disaster_Management_TestBed_Hands-on_Setting (accessed on 13 August 2018).
- Khaliq, K.A. Emergency Disaster Management. Available online: https://www.researchgate.net/publication/335146714_Configuration_of_’BATMAN_Adv’_for_Disaster_Management_TestBed_Hands-on_Setting (accessed on 13 August 2018).
- On the Roads in Germany—What Do I Need To Know? Available online: https://www.germanroadsafety.de/downloads/pdf/unterwegs_in_deutschland_en.pdf (accessed on 21 December 2018).
- Traffic Accident Fatalities Up by Roughly 13% in October 2018 Year On Year. Available online: https://www.destatis.de/EN/PressServices/Press/pr/2018/12/PE18_514_46241.html (accessed on 21 December 2018).
- Dekra Road Safety Report 2018—Transport of Goods. Available online: https://www.dekra.com/media/road-safety-report-2018-dekra.pdf (accessed on 14 December 2018).
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).