Problem Overview
Many traffic accidents in the United States are caused by fatigue driving. A study by the AAA Foundation for Traffic Safety estimated that around 328,000 crashes involving drowsy driving occur annually (Drivers are falling asleep behind the wheel, 2020). In 2015, an estimated 5,000 people died in fatigue driving crashes, according to a Governors Highway Safety Association report (Drivers and falling asleep behind the wheel, 2020).
Many drivers do not realize the dangers of getting behind the will while fatigued. According to the National Sleep Foundation, about 50% of adult drivers admit to driving while drowsy, and 20% admit to falling asleep behind the wheel at some point in the past year (Drivers are falling asleep behind the wheel, 2020). Driving while drowsy is similar to driving under the influence of alcohol and is characterized by the decrease of reaction times, awareness of hazards, and ability to sustain attention. Addressing the problem of drowsy driving is important to reduce the number of accidents and increase road safety.
Proposal Description
In order to develop a system that accurately determines whether a driver is tired or not, three kinds of hardware are combined. The first is a detection device (radar) installed around the vehicle that monitors the instability of driving. It tracks the driver’s ability to maintain a distance from the car in front of the vehicle and the ability to keep the vehicle steady in the middle of the road. According to the studies, fatigue decreases vigilance, situational awareness, and psychomotor and cognitive responses, which leads to driving performance becoming increasingly variable and unstable (European Road Safety Observatory, 2018).
The more complex the road geometry is, the harder it is for a fatigued driver to control the steering wheel and maintain a stable lane position (European Road Safety Observatory, 2018). Monitoring car movement allows the system to detect driving instability and determine whether the driver is fatigued.
The second type of hardware is an eye-tracking device in the car’s rearview mirror. It monitors the eye-blink rate to determine whether the driver is too tired to concentrate on the road conditions. According to the research by Haq & Hasan (2016), the person’s level of attentiveness is expressed by their eye-blinking rate. When a person is feeling fatigued or drowsy, their eye-blinking pattern changes. The less attention the driver pays on the road, the longer is the duration of eye blinks (European Road Safety Observatory, 2018). According to the study by Schmidt et al. (2018), the algorithm that detects blinks can be used to distinguish between alert and drowsy driver behavior.
The third type of hardware is installed in the steering wheel and monitors how many times the steering wheel is swung. According to the study by Li et al. (2017), when the driver is in a drowsy state, the frequency of his steering wheel corrections reduces markedly. The sensors mounted on the steering wheel monitor its movements and measure the fatigue state based on the frequency of minor steering corrections.
Purpose of the Project
The purpose of the project is to offer a service to detect whether a person is driving under fatigue. It aims to combine three types of hardware: a car movement detection device, an eye-tracking device, and a steering wheel movement detection system. The hardware collects the data that is analyzed by software that measures the fatigue state and determines whether a driver is tired or not.
Project Objectives
The goal is to have 1,000 cars of different models tested by January to determine the system’s effectiveness. The percentage of cars of each type should reflect the registered vehicles statistics. According to the Bureau of Transportation Statistics for 2012 (Number of U.S. aircraft, vehicles, vessels, and other conveyances, 2020), 183 million out of 254 million registered vehicles were “light duty vehicles with a short wheel base,” 50 million were “light duty vehicles with a ling wheel base.” Another 8 million vehicles were classified as vehicles with two axes and six or more tires, and around 2,5 million were classified as trucks (Number of U.S. aircraft, vehicles, vessels, and other conveyances, 2020).
According to the car demand statistics for 2020, crossovers are American’s favorite type of passenger vehicles, accounting for over 40% of automobile sales (U.S. car demand: By segment in April 2020, 2020). Different types and sizes of cars need to be included in the research, which require installation of different numbers of sensors.
Research Methods
Due to limitations in staff and funding, the data collected by the hardware will be analyzed using goal programming and regression methods. Regression analysis is a technique that investigates the relationship between dependent and independent variables. In this research, it will be used to establish the relations between the three types of data collected by the hardware and the accumulated level of fatigue. The goal programming method will be used to assess the criteria in the decision-making process. Personnel efficiency and the cost of hardware installation will be evaluated using both goal programming and regression techniques.
Physical Evidence
The hardware production process is allocated between the four departments: processing, production transfer, internal labor, and technical support. It starts with the internal labor and processing departments purchasing and receiving the parts: the radar, eye-tracking device, and steering wheel sensors. Then, they are moved to the assembly line, where the three parts are assembled into one. The technical support team designs and optimizes the software and installs it into the system, and the finished product is moved to the warehouse. The system is installed in 1,000 vehicles, and tested by the processing department. The data is collected, evaluated, and optimized in the internal labor department with the help of the technical support team. After the test is completed, the approved system is transferred to the next destination point.
The production process is organized based on the principles of lean manufacturing. They include value, the value stream, flow, pull, and perfection. The method argues that the manufacturer needs to precisely specify the product’s value, identify the value stream, make value flow without interruptions, let consumers pull value from the producer, and pursue perfection (Womack & Jones, 2005). The manufacturer needs to solve the customer’s problem completely by ensuring that the products work together, providing exactly what the customer wants, where, when, and if they want it, and continually aggregating solutions to improve (Womack & Jones, 2005).
The proposed driver fatigue monitoring system is intended to not only address the driver’s needs but provide a comprehensive way to improve the road situation, which should be taken into consideration during production. Testing is intended to ensure that the system is effective and satisfies all technical and customer requirements.
References
Drivers are falling asleep behind the wheel. (2020). National Safety Council. Web.
European Road Safety Observatory. (2018). Fatigue 2018. Web.
Haq, Z. A., & Hasan, Z. (2016). Eye-blink rate detection for fatigue determination. 2016 1st India International Conference on Information Processing. Web.
Li, Z., Li. S. E., Li, R., Cheng, B., & Shi, J. (2017). Online detection of driver fatigue using steering wheel angles for real driving conditions. Sensors, 17(3), 495. Web.
Number of U.S. aircraft, vehicles, vessels, and other conveyances. (2020). Bureau of Transportation Statistics. Web.
Schmidt, J., Laarousi, R., Stolzmann, W., & Karrer-Gauss, K. (2018). Eye blink detection for different driver states in conditionally automated driving and manual driving using EOG and a driver camera. Behavior Research Methods, 50, 1088–1101. Web.
U.S. car demand: By segment in April 2020. (2020). Statista. Web.
Womack, J., & Jones, D. (2005). Lean consumption. Harvard Business Review. Web.