Deep Reinforcement Learning-Based Pedestrian and Independent Vehicle Safety Fortification Using Intelligent Perception

Deep Reinforcement Learning-Based Pedestrian and Independent Vehicle Safety Fortification Using Intelligent Perception

Vijayakumar P., Jegatha Deborah L., Rajkumar S. C.
DOI: 10.4018/IJSSCI.291712
OnDemand:
(Individual Articles)
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

The Light Detection and Ranging (LiDAR) sensor is utilized to track each sensed obstructions at their respective locations with their relative distance, speed, and direction; such sensitive information forwards to the cloud server to predict the vehicle-hit, traffic congestion and road damages. Learn the behaviour of the state to produce an appropriate reward as the recommendation to avoid tragedy. Deep Reinforcement Learning and Q-network predict the complexity and uncertainty of the environment to generate optimal reward to states. Consequently, it activates automatic emergency braking and safe parking assistance to the vehicles. In addition, the proposed work provides safer transport for pedestrians and independent vehicles. Compared to the newer methods, the proposed system experimental results achieved 92.15% higher prediction rate accuracy. Finally, the proposed system saves many humans, animal lives from the vehicle hit, suggests drivers for rerouting to avoid unpredictable traffic, saves fuel consumption, and avoids carbon emission.
Article Preview
Top

1. Introduction

Emerging universal transportation has made people travel at their desired destination more complex and inconvenient because of the enormous vehicle population, unpredictable traffic on the road, and sudden accidents, creating the environment and peoples’ lives more difficult. An intelligent transportation system avoids such problems and ensures optimized traffic by giving valuable recommendations to the onboard vehicles to improve the safety throughout the journey for vehicles and pedestrians (Rajkumar S C & L. Jegatha Deborah 2017). Another critical reason for accidents is independent vehicles such as classical vehicles, bicycles, motorcycles, trucks, and others with no internet connection to create vehicle-type communication. However, even if this type of vehicle driver has smartphones for receiving emergency alerts, the driver side’s response rate is not efficient compared with the proposed system. Despite this, inattentive driving causes severe injuries to pedestrians as well as the environment. Moreover, drivers have fewer possibilities to take better solutions without intelligent recommendations before attempting the worsening state. Some of the following factors play a significant role in reaching worsens state:

  • Lack of concentration

Driver’s visibility is always worse at night than during the day, and in this case, an inattentive driver has a high chance of hitting the pedestrian on the highways.

  • Driving while intoxicated

If a driver consumes alcohol or illegal drugs while driving, their optic nerves are damaged and cause increasing the likelihood of an accident while driving.

  • Interruption

Many accidents occur due to distracted driving, and such are cell phone calls/texting, eating, drinking, and listening to music if a vehicle speeding on the highway, especially along with zebra crossings, chance to create an environment more dangerous and any distraction can exacerbate the causality for pedestrians and vehicles.

  • Poor light

Most street lights installed on highways illuminate only the pedestrian crossing place, although some people might cross outside the cross line at the low light side, resulting in a high-speed vehicle and distracted or careless drivers who can affect the precious life of a human or animal.

  • Damaged roads

Unexpected road damage caused by heavy rainfall, sudden road works, vehicle damage, and other factors causes the pedestrian or vehicle, resulting in tragedy.

The proposed system predictive results were obtained using intelligent agents such as Reinforcement Learning RL to solve unsafe transportation by providing alerts for independent vehicles and pedestrians. Another important key factor for avoiding an accident is to improve the responsive time from the server-side to avoid a collision, and it improves the results for getting a lower hit rate. Today, many vehicles have obstruction detection sensors used for vehicle parking, compact size, low cost, and good sensing range. The proposed system uses a Light Distance and Ranging (LiDAR) sensor, which is implemented on the roadside to calculate better detection results of pedestrians and independent vehicles, and the sensor flow is depicted in figure 1.

Figure 1.

Sensor Flow

IJSSCI.291712.f01

Complete Article List

Search this Journal:
Reset
Volume 16: 1 Issue (2024)
Volume 15: 1 Issue (2023)
Volume 14: 4 Issues (2022): 1 Released, 3 Forthcoming
Volume 13: 4 Issues (2021)
Volume 12: 4 Issues (2020)
Volume 11: 4 Issues (2019)
Volume 10: 4 Issues (2018)
Volume 9: 4 Issues (2017)
Volume 8: 4 Issues (2016)
Volume 7: 4 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing