Real-Time IoT-Based Connected Vehicle Infrastructure for Intelligent Transportation Safety

被引:11
作者
Sharma, Neerav [1 ]
Garg, Rahul D. [1 ]
机构
[1] Indian Inst Technol Roorkee, Geomat Engn Grp, Roorkee 247667, India
关键词
Animals; Roads; Urban areas; Real-time systems; Accidents; Safety; Cows; Computer vision; ADAS; transport safety; IoT; intelligent system; TRAFFIC CONGESTION; ACCIDENTS;
D O I
10.1109/TITS.2023.3263271
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
The transportation sector faces severe consequences due to the incrementing population influx yielding congestions, fatalities and haphazard traffic scenarios. Advanced Driver Assistance Systems (ADAS) assists highly in such scenarios by eradicating probable accidents and ensures traffic safety. This paper presents intelligent transportation systems (ITS) approach through the connected vehicle technology infrastructure. YOLO v4 (You Only Look Once) inspired real-time computer vision capable of detecting vehicles, pedestrians and animals at high efficiency (0.9777 mean average precision) is deployed on the GPU (Graphics Processing Unit) which offered higher frame rate of detection (74.26 fps). The locations of animals and potholes were mapped through consistent survey and mobile app which relayed the detected locations to the cloud server forming a geospatial database. Clustered locations from the geospatial database on dense transportation network were utilized for constructing animal and pothole hotspot zones. A basic level of display warning was triggered when the vehicle approached animal and pothole areas. Furthermore, advanced alert comprising of display and sound alert was trigger when the vehicle approached hotspot zones. This was implemented using real-time Internet of things (IoT) and cloud infrastructure applications for continuous vehicle's location monitoring and triggering as per the hotspot geo-locations. The proposed system ensured traffic safety and assisted in avoiding probable crashes and accidents that generally led to congestions and fatalities.
引用
收藏
页码:8339 / 8347
页数:9
相关论文
共 70 条
[1]  
Alpatov BA, 2018, MEDD C EMBED COMPUT, P120
[2]  
[Anonymous], 2021, Times of India
[3]  
[Anonymous], 2019, The Times of India
[4]   A Review on IEEE 802.11p for Intelligent Transportation Systems [J].
Arena, Fabio ;
Pau, Giovanni ;
Severino, Alessandro .
JOURNAL OF SENSOR AND ACTUATOR NETWORKS, 2020, 9 (02)
[5]   A real-time network-level traffic signal control methodology with partial connected vehicle information [J].
Bin Al Islam, S. M. A. ;
Hajbabaie, Ali ;
Aziz, H. M. Abdul .
TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, 2020, 121
[6]   Spatially resolved hourly traffic emission over megacity Delhi usingadvanced traffic flow data [J].
Biswal, Akash ;
Singh, Vikas ;
Malik, Leeza ;
Tiwari, Geetam ;
Ravindra, Khaiwal ;
Mor, Suman .
EARTH SYSTEM SCIENCE DATA, 2023, 15 (02) :661-680
[7]   Resource-Constrained Machine Learning for ADAS: A Systematic Review [J].
Borrego-Carazo, Juan ;
Castells-Rufas, David ;
Biempica, Ernesto ;
Carrabina, Jordi .
IEEE ACCESS, 2020, 8 :40573-40598
[8]   Machine Learning-based traffic prediction models for Intelligent Transportation Systems [J].
Boukerche, Azzedine ;
Wang, Jiahao .
COMPUTER NETWORKS, 2020, 181
[9]   Analysis of the Risk Factors Affecting the Severity of Traffic Accidents on Spanish Crosstown Roads: The Driver's Perspective [J].
Casado-Sanz, Natalia ;
Guirao, Begona ;
Attard, Maria .
SUSTAINABILITY, 2020, 12 (06)
[10]   Rural and urban traffic fatalities, vehicle miles, and population density [J].
Clark, DE ;
Cushing, BM .
ACCIDENT ANALYSIS AND PREVENTION, 2004, 36 (06) :967-972