Smart Road Accident Prevention System for Revenue Generation in Ghana Emmanuel Effah 1 , Mohammed Yussif Umaru 2 , Dodoo Frances Naa Densua 3 , Ghartey George 4 and Akogo Kennedy Kweku 5 1,2,3,4,5 Department of Computer Science and Engineering, University of Mines and Technology, Ghana 1 eeffah@wpi.edu, 2 myumaru@umat.edu.gh, 3 francesnaadensuadodoo398@gmail.com, 4 georgeghartey2@gmail.com, 5 akogokennedy@gmail.com Abstract—In this article, we present novel smart Internet- of-Things (IoT)-based driver alcohol and drowsiness detection for revenue generation (DADD4RG) testbed that can minimize road accidents and remotely report recalcitrant drivers with fines in Ghana. Road traffic accident (RTA) remains a major public health and development challenge in Ghana. Additionally, it is among the top 10 causes of death, draining 2.54% of Ghana’s gross domestic product annually of which, drunk driving and driver drowsiness constitute 40% of the causes of road accidents. This surge in road deaths and injuries has ignited demands for a sharper policy focus on road carnage. To date, the conventional 3Es (education, enforcement, and engineering) control measures have not achieved much due to a lack of the requisite intelligent road/traffic usage technologies desired to complement the engineering component of the 3Es control measures. There is demand for context-relevant solutions that are based on wider societal factors that impact road safety, which are currently lacking. The proposed DADD4RG testbed detects fatigued drivers and their drunkenness, send a notification to the driver/car owner, and log the associated fine of the offense including date, time, and location of the vehicle in the national vehicle usage database. A sample performance results of this testbed showed its contextual relevance and potential of minimizing fatigue-based road traffic accidents and improving national revenue generation. Index Terms—Driver Alcohol Detection, Drowsiness Detection, Revenue Generation, Road Traffic Accidents I. I NTRODUCTION The number of deadly traffic accidents keeps increasing every year, and the associated damages (mortality and socioe- conomic impacts) on societies is a major public health and developmental concern. Despite the prominence of Covid-19 disease in the current public health space, road crashes remain an important contributor to mortality. Recent data on global road crashes shows that more 1.4 million people die from road crashes globally, with most of these being the youth and people from developing countries [1]. In Ghana, more than 60% of road traffic fatalities occur in the economically active population (i.e., young persons under 35 years) [2]. The current daily statistics from Road Traffic Accidents (RTAs) data in Ghana show that over 72 persons out of every 100000 population suffer road crash-related bodily injury and over 8 persons of the same population die from RTAs [1, 2]. Besides the concomitant mortality and morbidity with road crashes, Ghanaian households spend an average of US$ 1687.65 in direct and indirect costs on severe injuries associated with road crashes, while many suffer considerable degrees of psycholog- ical distress [3]. To date, the conventional 3Es (education, enforcement, and engineering) control measures have not yielded the desired results due to a lack of the requisite context-relevant tech- nological measures on our roads. The reason being that best technologies in road safety management must be based on all the 3Es as well as the societal factors [1, 4]. This has mo- tivated the design of a novel, context-specific smart Internet- of-Things (IoT)-based driver alcohol and drowsiness detection for revenue generation (DADD4RG) testbed, which will not only prevent road accidents but also contribute to national revenue generation and driver discipline. Within this research arena, several state-of-the-art benchmarking solutions, such as referenced in [3, 4, 5, 6, 7], have emerged. However, these solutions are context-specific and so, do not meet all the desired performance and adoption requirements of most countries in Africa such as Ghana due to several technical challenges. • Existing solutions mostly present performance results on theoretical frameworks of either driver alcohol detection systems or drowsiness detection without real-world val- idation testbed or custom-built prototypes to bridge the gap between the existing theories and practices. • Current designs do not incorporate context-specific soci- etal or behavioral factors that affect road safety manage- ment. • There exist an urgent need for DADD4RG solutions that can only be acquired, deployed in vehicles without Wi- Fi coverage and managed by drivers without sufficient financial resources and technical expertise in IoT. The aforementioned technical challenges establish the need for an in-depth contextual evaluation of DADD4RG as well as a holistic account of the real-world experiences from design to the deployment of this novel technology that can be validated under both indoor and outdoor environmental conditions. This paper proposes a novel context-specific DADD4RG testbed that consists of the following: • A novel DADD4RG technology that integrate alcohol and International Journal of Computer Science and Information Security (IJCSIS), June 2023 https://google.academia.edu/JournalofComputerScience 1 https://sites.google.com/site/ijcsis/ ISSN 1947-5500