Journal of Engineering and Applied Sciences Technology

Introduction

Self-driving cars, also known as autonomous vehicles (AVs), represent a major shift in the transportation industry. In less than 10 years, the race to develop driverless cars, has grown to be one of the largest technological competitions in the world, with tens of billions of dollars invested in businesses and start-ups [1]. The ultimate goal is a consumer-driven, driverless, on-road vehicle; whether privately owned or a component of a centralised ride-sharing fleet, this is the area where the majority of investment has taken place. However, AVs have been used for much longer in other industries, such as mining, which face some of the same technical difficulties as on-road AVs, but not all of them. Despite the potential benefits, the development and deployment of self-driving cars face several challenges that must be addressed.

Literature Survey

Hancock et al., 2019; Sparrow & Howard, 2017, believes that driverless cars will represent the transportation of the future [2-3]. But there is some debate over the precise time frame for their possible arrival. Some authorities such as in also claims that driverless cars are already feasible and will be commercially available by 2020 [4-5]. Researchers in were more cautious and predict that it may take another two or three decades for driverless car technology to advance to the point where it is practical for widespread use [6-7]. The reason for the divergent timelines for the introduction of driverless cars can be attributed to differing perspectives on the magnitude of the technological obstacles that must be overcome for them to be safely integrated into the transportation system. The gradual automation of currently human-performed tasks while maintaining human supervision of the driving task is one popular theory about how driverless cars will replace human drivers on the road. According to this line of thinking and proposed a lane change assistance and automated motorway driving represent the next step, with Tesla's "autopilot" already making significant progress toward achieving this latter objective [8-10]. Authors such as and also conducted a research on cruise control and anti-skid braking systems, which are already widely used, represent the beginnings of autonomous driving [11-13].

Table 1: Some of the Literature Survey

Discussion
Key Technical Competencies-Hardware

A typical AV is made up of several essential hardware parts, including the physical platform, a collection of sensors, and on- board computational hardware.

Platform

The viability of various technological solutions to the autonomy problem depends on the type of AV platform. Larger vehicles can carry more onboard sensing and computing components, but they are typically heavier, harder to stop, and more damaging. Additionally, energy storage typically scales well with vehicle size, which is an important factor that can enable higher uptime percentages and the use of more power-intensive computing.

Sensor Technology

Sensors for AV platforms vary (Figure 1). Laser and lidar-based range sensors use reflectance data to detect lane markings and provide accurate long-distance range-to-object information. Rain and smoke can significantly reduce their capabilities. Modern cameras produce high-frame-rate, high-resolution images of the environment. These images have good dynamic range, showing detail in both bright and dark areas (Figure 2). Humans can drive well with visual sensing alone. If extracted, camera images contain more information than any other sensing modality. Cameras are more sensitive to day-night cycles than lidar, but they are cheaper and use less power.

Figure 1: Typical Sensor on the top of the car (Adapted from [14])

Computational Hardware

All onboard autonomy-related tasks, including scene understanding, navigation, and high-level control, are processed by computer hardware. Recent hardware trends have concentrated on power usage per computing unit to increase electric vehicle range. With systems like its Jetson AGX Xavi er that are power-efficient and extremely capable, Nvidia is a good example of a major player in this market.

Figure 2: Autonomous car view a) a range of camera views b) range scans (Adapted from [14])

Key Technical Competencies: Software Locating and Mapping

The foundations of AV operation include mapping and localization. There are various localization subtypes, and each one facilitates autonomy on a vehicle in a unique way. How can a robot move through an environment, creating a map of that environment, while simultaneously localising itself within that ever-changing map, has long been a major area of research in robotics. Using GPS or onboard localization systems, approximate localization—what you get on your phone's GPS—is typically used for broad route planning. Additionally important is relative localization, such as understanding that the car is currently 0.73 metres from the edge of the road. For safe vehicle planning and control, precise relative positioning with respect to moving objects, such as an oncoming car, is essential.

Planning, Decision-Making, and Control

Sensing and mapping are important, but how an AV uses that information to plan and act—whether to speed up, slow down, turn, or activate a turning indicator—is crucial to its survival. When braking is not possible, the planning system must plan safe actions like slowing down or abruptly changing lanes to avoid an unexpected obstacle. Accidents don't conflict with protecting people outside the vehicle, so they should be prioritised.

Challenges in the Adoption of Self-Driving Cars Real-World Testing vs. Simulation

AV developers struggle with the fact that self-driving cars are already safe—about one fatality per 100 million miles driven. Thus, gathering enough data from a few development vehicles under ideal conditions to prove a system's safety is difficult. Thus, autonomy developers use simulation. In high-fidelity simulation environments, researchers can simulate and evaluate specific weather conditions and pedestrian configurations at much higher throughputs than in the real world. Thus, much effort has gone into showing how well development can be used and transferred in simulation environments.

Corner-Case Issues

Corner cases are rare events that are difficult to predict, anticipate, and respond to. Self-driving cars struggle to handle these rare situations because their artificial intelligence systems cannot generalise like humans. More real-world data, billions of miles of driving simulations, and pathologically difficult scenarios are being collected to better handle these corner cases.

Weather and Climate

AVs struggle in a dynamic world. Day-night cycles, seasonal changes, and weather conditions like rain, snow, and fog change the environment's appearance and structure. These issues can be partially resolved using cutting-edge methods or sensors like lidar, which are less sensitive to appearance change.

Current State- AVs beyond the road

Avs have been used in mining, logistics, agriculture, and defence, among others.

A.Mining

Mining is large enough to support the financially intensive development of AV-related technology, its remote operation workflows are easier to automate, and there are fewer latency- critical scenarios, making it possible to occasionally hand over to a remote operator. Rio Tinto's AV haulage system recently hauled its one billionth tonne autonomously. Alternative technologies are needed underground because satellite-based GPS is unavailable.

B.Agriculture

Autonomous farming machines can sow, plant, and harvest crops. Despite dozens of AV trials, few long-term commercial deployments have slowed progress.

C.Defending

Modern defence theory assumes a complete blackout of communications and GPS-based positioning technologies (similar to underground autonomous mining trucks), leaving on-vehicle autonomy to make most decisions. These vehicles face mobility, perception, planning, and control challenges in densely forested and dusty, smoky urban environments. Finally, ethical issues surrounding autonomy in defence applications are gaining attention.

Conclusion

The development of self-driving cars represents a major shift in the transportation industry and has the potential to bring numerous benefits. However, the industry faces several technical, regulatory, and ethical challenges that must be addressed before widespread deployment can occur. To ensure the safe and responsible deployment of self-driving cars, it is crucial for the industry to invest in research and development, for policymakers to establish clear and comprehensive regulations, and for the scientific community to address the ethical and social implications of the technology [15-17].

References

1. S B Rajasekaran (2023) AI and Cybersecurity - How AI Augments Cybersecurity Posture of an Enterprise. Int J Intell Syst Appl Eng 11: 179-182.
2. P A Hancock, I Nourbakhsh, J Stewart (2019) On the future of transportation in an era of automated and autonomous Proc. Nat. Acad. Sci. USA 116: 7684-7691.
3. R Sparrow, M Howard (2017) When human beings are like drunk robots: Driverless vehicles, ethics, and the future of Transportation Research Part C: Emerging Technologies 80: 206-215.
4. J Hwang, W Li, L Stough, C Lee, K Turnbull (2020) A focus group study on the potential of autonomous vehicles as a viable transportation option: Perspectives from people with disabilities and public transit agencies. Transportation Research Part F: Traffic Psychology and Behaviour 70: 260- 274.
5. W Larson, W Zhao (2020) Self-driving cars and the city: Effects on sprawl, energy consumption, and housing Regional Science and Urban Economics 81: 103484.
6. P Bansal, K M Kockelman (2018) Are we ready to embrace connected and self-driving vehicles? A case study of Texans. Transportation 45: 641-675.
7. A Rasouli, J K Tsotsos (2020) Autonomous Vehicles That Interact with Pedestrians: A Survey of Theory and Practice. IEEE Transactions on Intelligent Transportation Systems 21: 900-918.
8. S Grigorescu, B Trasnea, T Cocias, G Macesanu (2020) A survey of deep learning techniques for autonomous Journal of Field Robotics 37: 362-386.
9. A Chehri, H T Mouftah (2020) Autonomous vehicles in the sustainable cities, the beginning of a green adventure. Sustainable Cities and Society 51:
10. J Stilgoe (2018) Machine learning, social learning and the governance of self-driving cars. Social Studies of Science 48: 25-56.
11. B Zheng, Z Hong, J Tang, M Han, J Chen, et (2023) A Comprehensive Method to Evaluate Ride Comfort of Autonomous Vehicles under Typical Braking Scenarios: Testing, Simulation and Analysis. Mathematics 11: 474.
12. S McLachlan, E Kyrimi, K Dube, N Fenton, B Schafer (2022) The Self-Driving Car: Crossroads at the Bleeding Edge of Artificial Intelligence and arXiv preprint 1-27.
13. D Adami, M O Ojo, S Giordano (2021) Design, Development and Evaluation of an Intelligent Animal Repelling System for Crop Protection Based on Embedded Edge-AI. IEEE Access 9: 132125-132139.
14. X Zhao, P Sun, Z Xu, H Min, H Yu (2020) Fusion of 3D LIDAR and Camera Data for Object Detection in Autonomous Vehicle IEEE Sensors Journal 20: 4901-4913.
15. A Kumar, R Krishnamurthi, A Nayyar, A Kr Luhach, M S Khan, et (2020) A novel Software-Defined Drone Network (SDDN)-based collision avoidance strategies for on-road traffic monitoring and management. Vehicular Communications 100313.
16. S Grigorescu, B Trasnea, T Cocias, G Macesanu (2019) A survey of deep learning techniques for autonomous Journal of Field Robotics doi: https://doi.org/10.1002/rob.21918.
17. H Ning, R Yin, A Ullah, F Shi (2021) A survey on hybrid human-artificial intelligence for autonomous driving. IEEE Transactions on Intelligent Transportation Systems 1-16.