Table of Content
Technological advancements have created the need to develop more efficient vehicles in the automobile industry. The development of automated vehicles set the basis for self-driving or autonomous vehicles. Consequently, automobile and technology companies have embarked on numerous collaborations to create driverless (autonomous) vehicles. Research has mainly focused on the technologies and software used to develop these vehicles. Majority of the literature also indicate the benefits and setbacks of autonomous vehicles. This paper provides an overview of the major issues associated with autonomous vehicles. It also focuses on the major companies involved in the development of driverless vehicles and test drives. It also provides the recommendation of a prototype to improve autonomous vehicle technology. It concludes by identifying key areas of development for future autonomous vehicle technologies.
Modern technology has resulted in many possibilities in all sectors of the economy, including the transport sector. Modern automotive technology seeks to design and create vehicles that promote greater efficiency, speed and safety. The technology has led to the development of autonomous or self-driven vehicles (McCall et al. 293). Despite the development of automated cars, new approaches see to create self-driving or robotic cars to improve automobile technology. Self-driving cars are vehicles that have the capacity to sense the environment and navigate through tracks without human input. Various techniques are used detect surroundings. These include: GPS, computer vision, odometry, lidar, and radar. They also use advanced control systems that interpret sensory information and identify the necessary navigation paths. Self-driving cars also have technologies that analyze sensory data in order to distinguish different cars. This helps to plan the most suitable path towards the desired destination. Some of the main companies that manufacture self-driving cars include Google, Tesla Motors and Volkswagen. Self-driving cars are important to promote efficiency, reduce road accidents and expand the motor vehicle industry. However, there is need to incorporate new and efficient technology to reduce the inefficiencies and accidents associated with autonomous vehicles.
Within the next decade, there will be an increase in the number of autonomous vehicles, especially with the amplified level of technology applied in their development. Development of the vehicles is being informed by increased research on this area to establish the limitations that should be tackled to make the cars more functional and beneficial. It is also critical to look deeper into the social, technical, and policy issues surrounding the design and use of autonomous vehicle and the related technologies (Look & Shrobe 277). Evidently, the functionality of the vehicle depends on the ability of the vehicles to use computing capabilities to communicate with other vehicles and with the infrastructure. The communication will allow better navigation of the vehicles. While the modern vehicles are already able to communicate, there is no limitation to how far technology can develop. With more advanced research and prototyping, it will be possible to come up with even more intelligent designs for the autonomous vehicles.
Research has already been carried out on many aspects of the autonomous vehicles, including their history, the current models, techniques used in the navigation of the cars, and the tasks involved in the working of the vehicles. However, this is not enough since there are two main areas that should be the focus of future investigation on the development and use of the autonomous vehicles (McCall et al. 193). One of the areas, and the focus of the current study is the improvement of the modern cars through prototyping. In this case, the idea is for more research to be carried out in the ways of improving on the current design of the vehicles to make them more efficient, faster, and safer. The second area is more research on addressing the drawbacks and limitations in the current vehicular system (Look & Shrobe 277). Research on the current working of the vehicles is critical as it identifies the gaps that should be filled and questions that should be answered in the future research efforts.
Self-driving (Autonomous) Vehicles
Just as the name implies, the self-driving cars are operated without involvement of a person as a driver. The cars have been in the process of being developed, such that they can operate by sensing their environment. Self-driving cars mainly operate from the central processing unit that coordinates key technological functions of the car. They are electrically designed and programmed to operate effectively. However, the world is yet to witness a fully autonomous car running on the road. These kinds of cars have been developed already, but those permitted to operate on the road are only partially autonomous. The ones allowed on the road have the drivers on the seat who is prepared to take control of the car if need be. However, there is still hope in the near future to have vehicles on the road that are completely autonomous (McCall et al. 193). After all, technology is still developing with an increase in research projects that are geared towards this end.
It is plausible to note that autonomous cars do not operate as the automatic ones. Thus, it is necessary not to confuse the automatic cars for autonomous ones, like in the subject of the current discussion. Autonomous suggest the potential for independence in operation without human mediation (Bimbraw 191). Most of the current projects that were geared towards autonomy only managed to design vehicles with automatic capabilities. Therefore, this means that the automatic cars heavily depend on the simulated indicator of the environment within which they operate. On the other hand, autonomy suggests the lack of such indicators and the ability of the design to sense the environment, amid uncertainty, and operate without the input of a human driver. The ability to compute the systems and use computerized sensing capabilities highlight the potential of the autonomous vehicles (Look & Shrobe 278). The vehicle might use communication capacities of information systems, but without the assisted operation from a driver.
Companies are already producing vehicles that are partly automated, and there is a growing competition to bring into the market more sophisticated vehicles into the market. There are cars with the capacity to provide a high degree of automation, but in certain situations, especially when driving on a highway) (Driggs-Campbell, Shia & Bajcsy 59). These vehicles are already in the market, with relentless efforts to make them even more sophisticated, towards the end of producing completely autonomous vehicles. The role of the driver is completely changing with the development of these sophisticated designs of cars. The role of the driver in the manual and completely autonomous vehicle is clear-cut (Bimbraw 192). The role of the driver in the autonomous vehicle has been replaced with computers that are programmed to navigate the vehicle devoid of the input of the driver. One of the common views is where the driver is seen as taking the place of the passenger, where no involvement is initiated in driving the car.
It is expected that the fully autonomous cars will arrive into the market in the coming years, although the transition from the automatic might be a cause for concern. Although testing has been done, there is no certain way of indicating whether the vehicles can actually perform with the level of efficiency expected in the modern transport system. There are concerns about bringing cars into the market that are incomplete and imperfect, necessitating the input of the driver to play the oversight role, and at times control the vehicle. However, the input will depend on the design process of the vehicle (McCall et al. 194). Research appears to point to the potential of developing a completely self-driving car, which will travel completely without the driver. If that time comes, the cars will provide greater possibility in terms of speed, efficiency, and safety.
While the cars might have entered the market only recently, the idea behind their use dates back to decades ago. Precursory to the driverless cars and some systems, demonstrating their capability has been traced back to the 1920s and 1930s. In 1956, the Electri Utility Company brought to the fore the idea of “electricity being the future driver.” The idea became the precursor to the development of the actual autonomous car. However, it was until five decades later, in the 1980s, that the actual autonomous car was developed. Precisely, the car was established in 1984 under the Navlab and ALV projects at Carnegie Mellon University (Look & Shrobe 279). The second one emerged in 1987 under the Mercedes-Benz and the Eureka Prometheus Project at Bundeswehr University Munich (Liu et al. n.p). In fact, since the two vehicles came into being, there have been many other firms and research companies that have come up with workable prototypes for autonomous vehicles.
From the 1987’s VaMoRs, major strides were made in the efforts to design a prototype based on an actual autonomous model. Ernst Dickmann, a German engineer, made a step further in developing the autonomous technology after fitting into a sedan bank of cameras, together with 60 micro-processing systems geared towards detecting items on the road. These devices were fitted both at the back and at the front of the vehicle. The model presented a major innovation as a vibrant vision towards the development of actual self-driving car (Kumar et al. 259). The idea is to come up with a vehicle with imaging capabilities for filtering inappropriate “noise” as well as focusing solely on the relevant objects in the roads. The imaging technique has become a critical part in the development of the actual self-driving vehicles, based on the identification of possible hazards and knowing where they are located. The German model was so effective that it navigated famed Autobahn at a speed of up to 60 miles/hour.
Another breakthrough came in 1995 with the design of the General Atomics MQ-1 Predator. Commonly, the self-driving car is considered a milestone, especially from the point of view of changing the human into a passenger and not the driver. However, there is another way of viewing the technology, as having vehicles that are created to move on their own and completely autonomous. The idea borrows from the model of the drones, especially the General Atomics Predator, which is one of the commonly known technologies that would assist in developing the autonomous vehicles. The model has remained the pilot for drones designed globally for about two decades. The drones are commonly used in the military arena, where they are not only military vehicles, but also assist in the defense and intelligence purposes. Technologies are the basis for the development of the model, including radar that is used as the navigation aid. With the navigation aids that are programmed into the model, it is possible to move through clouds and smoke (McCall et al. 196). They are also fitted with thermal imaging cameras, which make it possible for the vehicles to make out their path even in darkness. The success of these prototypes indicated the potential for large-scale development and use of these cars that are driven by technology. Research has pointed to this possibility in a very near future.
There are important features that fit into the car allowing for the self-driving capability. In Figure 1, the RADAR is usually used by the cars in performing detection of other vehicles that come close to the car through detection of their range, velocity, and angle using the radio waves. The feature is also useful in the detection of the weather formations. LiDAR (Light Detection and Ranging) is popularly used in the autonomous cars for the purpose of affording the vehicles high-resolution maps (Kumar et al. 261). The maps allow for effective navigation and determine the distance between vehicles by using a laser light to illuminate the target. Global Positioning System (GPS) provides the sensory capability of the vehicle. The cars are fitted with a sensor chip that offers the information on the location and the weather conditions all over or on close by earth. With the help of motion sensors, Odometry is used by the cars in providing the approximated change in the position of the car over time. The cars have Computer Vision, which is used in aiding navigation for the completely autonomous vehicles. The uses of this feature include the generation of the map showing the environment of the car and detection of any obstacles in its path. It can also be useful in the detection of some task-specific incidents.
The cars do not depend on the input of a person to navigate through the surrounding as would be the case using the human driven cars. Therefore, the technology underlies the mechanism that allows these cars to move around without human efforts. The cars are fitted with sensory capability, allowing them to sense their environment and move around with ease (Kumar et al. 261). They also have advanced control systems with the capability for interpreting the sensory information for the purpose of identifying the suitable navigation routes and the possibility for hindrances and pertinent signage. The cars are designed in such a manner that they are able to update the navigation information on the basis of the input through their sensory capability, making it possible for the car to recognize their positioning even in case of change in the external conditions or upon entering any uncharted environment (Bimbraw 191). The sensory information is analyzed using a control system. The analysis does not only point to the direction and any obstacles, but also help in recognizing any other car using the same road, allowing the correct establishment of the most feasible path to reach where the car is destined.
Laser Illuminating Detection and Ranging (LIDAR). LIDAR is a software technology that builds a 3D map for the car (Ulrich). This technology also allows the car to view potential hazards on the road and bounces a surrounding laser beam across all the surfaces of the car. This allows the car to determine the profile of the object and its distance. This technology is mainly used by the Google driverless Car. Radar is also a technology used in autonomous vehicles. Radar units are packed with sensors on the car at the front and back. Radar units mainly allow the car to avoid any impact by sending signals to an on-board processor. This alerts the system to apply brakes or move out of the way whenever applicable. This technology works collaboratively with gyroscopes, wheel encoder and inertial measurement units. High powered cameras are also used by autonomous vehicles to act equivalent to the human eye (see figure 2 for the camera image). Each vehicle camera has a 50-degree field of view which is accurate to almost 30 meters (Blissing, Bruzelius, and Eriksson 4). Sonar technology is also used to enable the car to cross-reference data across different systems in real time. Map systems such as Google Maps and GPS data are also used to track the course of the vehicle and provide necessary directions.
Wheel odometry is the most used technology in estimating the positioning of the autonomous car. The technique is utilized in estimating the position through the computation of the wheel’s revolutions when it comes into contact with the road. Accurate translation of the revolutions can be achieved through linear displacement in relation to the road (Bimbraw 191). In this case, the wheel rotation is measured using encoders as showed in figure 3 below.
The technique uses relative positioning method. However, the technique has a drawback in that it is affected by position drift and a lack of accuracy due to slippage of the wheel, leading to accumulated errors over time. Orientation and translation mistakes in the method augment proportional to the overall distance traveled. However, the method remains common because of the ease in its application and the fact that it is less costly (Bimbraw 191). It also allows for high rates of sampling and reflects a high degree of short-term accuracy.
The vehicles classified as Level 0 are those that use the driver to control them at all times. Level 1 vehicles use the vehicular control like the electronic control of stability or automated braking system. Level 2 vehicles normally use two controls which are concorded like the one used for lane-keeping and adaptive cruise control. Those belonging to Level 3 use the vehicular safety control systems, which are completely conceded. However, the driver on certain situations normally does the process (Kumar et al. 265). The Level 4 vehicles normally have all the safety-critical functionalities that are carried out with the vehicle from the start to the conclusion and comprise unoccupied vehicles.
Tasks in Autonomous Driving
Research on technology has concluded that autonomous driving is an extremely complicated system consisting of a number of diverse tasks.
The figure 4 above indicates the tasks involved in driving an autonomous car, a process that involves three phases, sensing, perception, and decision. The idea is that to be able to achieve operation of an autonomous car in an urban setting, where the traffic is predictable, there must be effective interoperability of a number of real-time systems. The processes include “sensor processing, perception, localization, planning, and control.” It is critical to note that the prevailing effective implementation of self-driving is typically LiDAR-based (Driggs-Campbell, Shia & Bajcsy 59). In fact, this is primarily because of the reliance on the LiDAR for the mapping, localization, and avoidance of obstacle. The other sensors are utilized for functions that are peripheral.
Sensing. The car is usually fitted with a number of sensors, which collect the data on the environment within which the vehicle will travel. The cars are fitted with more than one sensor to enhance safety and reliability. Some of the commonly used sensors include GPS and Inertial Measurement Unit (IMU) (Liu et al. n.p). The system plays the role of localizing the vehicle by sending a report on global position together with inertial updates. LiDAR is the other kind of sensor, which is used in mapping, localizing and avoiding obstacles. It bounces laser beams from the surface and measures the refection time as a way of determining the distance. The camera is used for recognizing objects that are on the road being travelled (Agatz n.p). It also helps in tracking tasks, including discovery of the lane, recognition of traffic light, and detection of the pedestrian, including such other tasks. Radar and Sonar are normally the final line of defense in avoiding obstacles (Driggs-Campbell, Shia & Bajcsy 60). The data they generate is used in indicating the distance from the closest obstacle. The vehicle applies brakes in the event of identification of an object that pose a risk of collision.
Perception. After obtaining the data from the sensory function, the data is normally fed on the perception task of the vehicle. Perception relates to the task of understanding the environment of the vehicle. There are three primary tasks in the function of perception, localization, detection of objects, and their tracking. Localization refers to a sensor-fusion task, like GPS/IMU and LiDAR, whose data can be applied in generating ground map on high resolution and infrared reflectance. Detection of objects within the route that the vehicle is travelling is achieved in the event of real-time localization (Liu et al. n.p). The task is also more effective in urban settings. Development of vision-based Deep Learning technologies have allowed for better detection of objects by the autonomous vehicles. The technology is also useful in tracking of the object. One type of Deep Neural Network is the Convolution Neural Network (CNN), which is useful in the role of recognizing objects (Driggs-Campbell, Shia & Bajcsy 60). Object tracking is a function used in tracking a moving object by following its trajectory. After applying the recognition function, the tracking can be performed to keep track of other moving cars or even a person crossing the road. Indeed, this plays an important role in preventing an incidence of collision with another car or hitting a crossing pedestrian. Deep learning methods have developed in the recent past to perform accurate object tracking.
Decision. The understanding of the environment of the vehicle leads to the decision function of the vehicle. This is the point during which the action plan is generated accurately and in real time. Probabilistic processes and Markov chains are the most common tasks under decision-making. When a person is driving a car, among the greatest challenges is predicting the possible action that will be taken by the other driver in the traffic (Driggs-Campbell, Shia & Bajcsy 60). The problem is especially witnessed in roads with more than one lane or at change points in traffic. Prediction in decision-making is critical to safety in such situations. Autonomous vehicles are fitted with the mechanism to make accurate predictions. Path Planning is a complex action in dynamic environments, particularly in the event that the vehicle has to apply its complete maneuvering abilities (Liu et al. n.p). A more unsophisticated strategy would be searching all potential routes using a cost function in the determination of the most suitable course. However, the approach would need a great deal of computational resources and might experience some errors (Agatz n.p). To address the challenge, there has been the development of algorithms, which are used by probabilistic planners. There are two main mechanisms in the avoidance of obstacles when using autonomous vehicles. The first approach uses traffic predictions, while the other one uses radar data.
Working of Google’s Cars
In Figure 5 below, Google’s development in autonomous car is clearly depicted. Indeed, to understand the working of the autonomous cars, Google’s cars can be used as a case study. When preparing to move the car, the “driver” begins by setting the destination. The software fitted in the car computes the path and begins the car. The LIDAH, located on the roof of the car performs the sensing function through rotating at a 60-meter range. The data obtained is used in creating a dynamic 3-D map showing the present environment of the car (McCall et al. 197). On the left wheel, there is a sensor which is used in monitoring the movement on the side of the car for the purpose of detecting its position in relation to the map. In the front and rear bumpers, the car is usually fitted with radar systems which are used in calculating the distance from the obstacle. The car has artificial intelligence (AI) application that has a connection to the sensors and Google Street View’s input. It also connects to video cameras within the vehicle. The AI plays the role of simulating the perceptive and decision-making capability of humans. Thus, it plays the roles of steering and applying breaks as would be the case for humans in a manual car. The software consults Google Maps for notifications in case more information about landmarks, traffic lights, and signs are required (Bimbraw 195). The car has an override task for allowing immediate control by a person in case of such a need.
Figure 6 shows a model of autonomous car designed by Tesla Motors. Major companies have established projects to create autonomous vehicles. As of 2016, over 33 corporations have expressed their interest in developing autonomous vehicles. These companies include motor vehicle manufacturers such as Toyota, Volkswagen, Volvo and Tesla Motors. Technology companies such as Google and Baidu also collaborate with motor vehicle companies to create self-driving vehicles. The Google self-driving car is one of the pioneer technologies in driverless technology. The project was launched in 2014 with over 5 prototypes so far. French PSA Groupe inclusive of DS, Citreon, and Peugeot has projected that their autonomous vehicles will be available to the market by 2021 (Ulrich). Tesla Motors has already launched its autonomous vehicles on its Model S range of cars across the US. Honda also received approval in California to test autonomous vehicles. ADAS is also a self-autonomous vehicle launched by Honda. Mercedes also launched its F015 autonomous vehicle project; its vehicles will be ready in 15 years (Ulrich). In 2015, Ford launched its Smart Mobility plan to manufacture quality autonomous vehicles in the next decade. Nissan/Renault has also announced an ambitious autonomous car project to sell 10 cars by 2020. Delphi, one of the major automotive suppliers has also created a network of sensors and software for self-driving cars. In 2014, BMW and Baidu partnered to create a semi-autonomous prototype by 2015. In 2016, DAF, IVECO, MAN. Scania, Volvo, and Daimler collaborated to complete their semi-autonomous convoy of trucks. However, these trucks still require human drivers as a precaution
Major strides have been made globally in terms of designing driverless cars (see figure 7 below). Major tests are conducted across different transport systems in the United States, Europe and Asia. Cities in countries such as France, Italy, United Kingdom and Belgium have established plans to operate relevant transport systems for self-driving cars. Countries such as Germany, Spain and the Netherlands have already permitted autonomous car tests in public traffic. In 2015, the United Kingdom launched major public trials for LUTZ Pathfinder in Milton Keynes (Ulrich). In France, the Peugeot-Citroen conducted its first trials in public roads across Paris. The vehicle navigated almost 136 miles (300km) without supervision across authorized stretches of roads in France. In 2016, the Peugeot-Citroen trials were extended to cities such as Strasbourg and Bordeaux. French companies such as Valeo and THALES have also conducted tests on the driver less car system across Europe. Major companies such as Google and Tesla Motors have also been involved in these road testing projects. As of August 2012, Google had completed over 500,000km (300,000) autonomous driving miles. These tests were accident free and involved over 12 cars on the road. The company also tests single drivers in pairs. In 2014, Google further launched a prototype of its driver less car. By 2016, Google has test driven its range of driverless cars in autonomous mode totaling to 1,500,000 miles (2.4 million kilometers). In 2015, BMW and Baidu tested their semi-autonomous modified 3-Series BMW across an 8.6 mile route in Beijing, China.
The self-driven car has come to be a reality across the world and might become the future of driving in the transport sector. Autonomous vehicles hold a promise for their availability in the market, especially if they assume the same pattern as other technologies within the automobile industry (see table 1).
Figure 8 indicates that if the autonomous vehicles follow the path of other technologies in automobile sector (see table 1), then the autonomous vehicles will dominate the sales of vehicles in the next one to three decades. It will take two to four decades to dominate travel, with the potential for market saturation.
Computers are speedily taking over the world of driving. However, the reality of the cars taking over the roads is yet to come through. Together with the design of driverless car designed by Google, the commercial availability of these cars is yet to take place. In fact, this is despite the fact that Google has logged thousands of hours in the United States. The world of technology has availed different kinds of designs of autonomous cars. They are founded on GPS sensing information to assist in navigation. The level of safety offered by the cars is the basis for the need to create them on a large scale to save the world from the increased level of traffic accidents (Agatz n.p). If successfully tested, the vehicles will be easily accepted in the market because of the promise of safety and efficiency.
The vehicles have the ability to utilize various technologies, where the car displays information, and hence allow the movement in innovative and novel ways. However, there are some suggestions that the production of the cars could lead to issues with the current traffic and insurance regulations utilized in the cars that are driven by humans. Under those premises, a great deal of research is being carried out in the United States and Europe and also in other regions of the world. Various players in the industry have indicated that it is a matter of time before the daily commuting of the people is outsourced through technology. Technology improves every day with increasing computational power of software and hardware (Liu et al. n.p). It s worth noting that those industries that capitalize on technology benefit from the improved application. In fact, the same dynamics are expected within the transport industry.
Advantages of Autonomous Cars
Automated cars offer numerous benefits to users and transport companies. The potential reduction of traffic accidents and collisions is one of the major advantages of autonomous cars. Figure 9 indicates a preliminary model showing the changes in personal risk over time following implementation and use of autonomous cars.
Traffic collisions often lead to massive deaths and injuries. In most cases, survivors develop different disabilities, which end up affecting their quality of life. These accidents are also costly for individuals, vehicle owners, and motor companies. For the individuals, loss of life and the impact on their health, including the cost of treatment have serious impacts. For the owners, the impact involves loss of property and costly lawsuits some of which end up in hefty compensations (Agatz n.p). Motor companies are also affected in the loss of property and damage to their reputation. Hence, reduction of such collisions in the use of the autonomous cars will lead to massive savings for individuals, owners of the vehicles, as well as motor companies.
Evidence from research indicates that human-driven errors are the leading causes of traffic accidents. Some of these errors include; tailgating, rubbernecking, and delayed action time. Some drivers also cause accidents due to destructions such as cellphone use and aggressive driving. Evidently, all these are errors, and hence, accidents that can be avoided. Autonomous cars have the capacity to navigate road systems and reduce the effects of human error given the reality that they do not rely on the input of humans to navigate through their environment. Self-driven cars also operate on advanced machine systems and software that reduce the likelihood of accidents (Wang and Nakao 32). The computational capabilities and use of deep learning algorithms in navigating the cars allow for navigation with a potential for high level of accuracy. For instance, the vehicles are able to detect and track objects on their route, hence avoiding obstacles that could otherwise cause accidents (de Paula Veronese et al. 520). A study provided estimates that application of autonomous vehicles in the United States could prevent 90 percent of the accidents related to automobiles, and save the country billions of dollars.
Autonomous cars offer the benefit of performance. Compared to the manned cars, the autonomous cars rate higher in terms of performance, especially under those circumstances that are characterized by uncertainty. As long as the car is secured on the road using the computerized system, it will move along the right path to the destination (Liu et al. n.p). Therefore, the cars can be able to travel even under circumstances that could pose a challenge to humans, such as in the darkness or cloudy environments. In fact, applying the capabilities of the autonomous cars, it is possible to navigate in an environment that would otherwise result in an accident for the human-driven cars. They have the potential to move at a very high speed, allowing them to arrive fast and without causing accidents on the road ((Bimbraw 197). They have also been shown to provide a smoother ride that would not be possible when manned vehicles are used.
The unnecessary costs are eliminated after there is a reduction of the number of road accidents taking place in the country. The use of self-driven vehicles points to a reduction in the labor cost because of the capacity to operate without a driver. The cars will also relieve travelers of the navigation and driving tasks. It will save travelers energy because they can simply sit and enjoy the ride without their input. The autonomous cars will save the burden of the driver and the many prohibitions, including the use of the mobile phone while driving. The travelers will also be able to reach their destination faster because of the enhanced speed and the reduction in traffic congestion (Bimbraw 197). There is the potential for increased road capacity by using the self-driven cars. The capability of communication between vehicles will also play the role of adding to the current road capacity. Effective control of traffic due to the ability to communicate will relieve the need for traffic police and other manual traffic controls.
There are possible concerns about possible theft of the autonomous car based on the fact that they are self-driven. However, those issues have been addressed in the computerized design adopted by the car. In case the car has been stolen, the details of the incidence are provided using the sensors fitted into its design (de Paula Veronese et al. 522). The information is used for ease in tracking the car, hence reducing the incidences of car theft. Saving on the stealing of vehicles will have greater impact on crime in the country and save vehicle owners and motor company from the loss of property. The element will also reduce the need for insurance covers for stolen or lost vehicles, money that can be channeled to other beneficial activities such researching based on making the technology even more advanced.
Carpooling is a concept that has gained interest in the recent past because of the problem of carbon emissions and climate change. In the case of carpooling, use of the self-driven cars has been associated with a decrease in the overall number cars traveling on the road at a given time. In fact, this is a critical factor behind the reduction in pollution (Liu et al. n.p). The cars are also more advanced regarding fuel efficiency. In addition, the reduction in fuel use is evidenced by the efficiency in the control of traffic. Use of the vehicles has the potential to free the space that would otherwise have been used for parking. The land can be put to more beneficial use such as building parks, housing, and retail outlets among others. The cars also support novel business models like mobility as a service. The autonomous cars will be a source of greater levels of mobility for different groups even those who are not able to drive, including children, the elderly, and individuals with disabilities.
Drawbacks of Autonomous Cars
Despite the numerous benefits offered by self-driving cars, there are also multiple limitations. The limitations have to be addressed in the future designs to make the vehicles more functional, efficient, and safe. At levels 3 and 4, the car must be in the position to sense the surrounding activities with the use of the various sensory techniques, including GPS, LiDAR, IMU, cameras. Therefore, based on the sensory input, the car will need to be localized and be in a position to make real-time decisions on the navigation route within the particular environment. Given the vastness of the data obtained through the sensing process as well as the increased complexity within the computing pipeline, driving the autonomous car creates a challenge in relation to the power of computing and the use of electrical power (Pozna and Antonya, 14). The current designs necessitate equipment of the cars with more than a single server. In addition, each of the servers should have multiple high-end CPUs as well as GPUs. The design is critical but presents a number of challenges for the cars.
Self-driving cars incorporate relatively new technologies that are complex and expensive. Implementing the new models necessitates a great deal of investment, with the cost trickling down to the customers. Hence, it becomes challenging for the general public to afford the cars. Besides the challenge in the use of computing power, there is a challenge in terms of the amount of electric power that is used to drive the autonomous car. Power supply as well as heat indulgence is a challenge in the design of the vehicles (de Paula Veronese et al. 522). A typical car is revealed to consume thousands of Watts, and as a result, places high demand on the power system of the vehicle.
New challenges arise with the current transport systems and the type of transport system required for the navigation of self-driving cars. It will be difficult to coordinate the navigation of manual cars and self-driven cars. Sometimes car technology may fail leading to accidents. Whereas they are fitted the best technology and software, self-driving cars have difficulty in interpreting human signals. These cars also lack the capacity to maintain safety across all-weather conditions. For instance, heavy rain can damage the laser sensor mounted on the roof of the car. If the system is used by an inexperienced driver, technological failure can led to accidents. Self-driving car technology requires drivers to be educated on the car’s information and control systems. For instance, some cars may experience undetectable errors when updating programs leading to machine failure (Ulrich).
Federal and state licensing infrastructure also limit the use of self-driving cars. The government has the responsibility to guarantee the safety of citizens. However, the development of autonomous creates challenges in defining safety and the most efficient use of autonomous vehicles among users. As of 2013, only four states in the US had legalized the use of autonomous vehicles. Despite their legalization, they have to go under stringent regulatory conditions across all these states. Local car licensing offices must ensure the vehicles operate efficiently and safely as advertised by manufacturers. Currently, the power, computer, engineering requirements, sensors and software account for almost S100, 000 (Casner, Hutchins and Don 73). Ethical issues such as loss of privacy and confidentiality also emerge from the use of self-driving cars. In order for computer software to operate the vehicle, owners must submit personal information to guarantee safety, security and proper functioning. In some cases, the car may identify the location of the driver.
Therefore, self-driving cars create a wide array of security issues for users. Currently, self-driving cars are scarce; consequently they attract high end users who are key targets for hackers. Due to their high dependence on software technology, autonomous car systems can easily be hacked. Users are also concerned with the potential of the computer to collect personal data such as credit card information, social security numbers and medical history. However, such information is prone to privacy breaches and hacking. Environmentalists are also concerned that self-driving cars will cause greater pollution. These autonomous cars require multiple technologies and electric emissions that consume more energy compared to regular cars. Economically, self-driving cars may cause massive unemployment. The transport sector is provides numerous employment opportunities to public transport drivers such as taxi and bus drivers. With autonomous transportation, majority of these drivers may lose their jobs.
Similarly, accidents and crashes observed in autonomous vehicles are fatal. For instance, in 2016, a self-driving Tesla Model S was involved in a fatal accident that killed one occupant (Caceras-Cruz, Arias, Guimarans and Juan 32). The crash occurred in Florida when the car was in autopilot mode and crashed into an 18-wheel tractor trailer. Formal investigations into the accident by the National Highway Traffic Authority indicate the car did not sense the 18-wheeler at an intersection on the highway. The car failed to apply its breaks leading to the crash. After passing under the trailer of the truck, the autonomous car continued to travel. This highlights concerns that autonomous cars may fail to sense impromptu issues emerging across highways. The NHTSA further commissioned a report from Tesla Motors on over 25,000 Model S Tesla self-driving cars (Ulrich). Tesla autonomous cars users have also complained that the vehicles have numerous loopholes. In one case, a driver stated that his car drove straight into a guardrail at a speed of 80 miles/hour. Google driverless cars have also been involved in numerous accidents. Google accident reports indicate that Google driverless cars were involved in 14 collisions where other manual drivers were at fault 13 times. However, in 2016, the Google software in their driverless cars caused its first major accident (Ulrich). As of 2015, Google CEO Sergey Brin confirmed that Google driverless cars were involved in 12 collisions (Ulrich). Eight of these crashes were caused by the cars when they were rear-ended at a traffic light or stop sign. Three were caused by other drivers and one caused by a Google employee who was manually driving the car. In February 2016, a Google driver less car was involved in an accident when it attempted to avoid sandbags that blocked its path. During this maneuver the car struck a bus.
Based on the challenges experienced by companies to create safer and more efficient autonomous cars, this research proposes the integration new testing software technologies. This new prototype will have high quality sensor equipment that detects different types of movements such as body movements. However, this prototype is fully autonomous. This means that it will run without a steering wheel, a brake pedal, brake pedal and steering wheel. The car operates with the push of a button. In most cases, autonomous vehicles continue to operate even after an accident injures the user. However, this prototype has an emergency system that can detect when the occupants are incapacitated or injured. This is because the sensor equipment can sense the body movements and heartbeats of the car’s users. A slow or faster heartbeat will indicate to the sensor software that the user is injured. When the heartbeat stops, the software can alert the manufacturer or traffic police. The software is also internet based whereby users can use the emergency button as a warning system when the car malfunctions. This emergency will notify the manufacturer who will notify traffic police. The prototype’s in-built sensor equipment can also detect body movement and impending vehicles across the highway. The high quality sensors can detect information 300 miles away.
This new prototype will also detect pedestrians and cyclists on the road as they cross. The sensor technology is also designed to sense traffic lights across urban areas that indicate pedestrian crossing. These sensing capabilities are designed to identify the different types of lights used in traffic equipment and emergency stop signs. The prototype also integrates notification equipment. Based on the different lights and body movements recognized by the sensor, the user is notified. The sensors will either identify a heartbeat of feet movement across the road. The software will help the car identify body movements and reduce the high rates of pedestrian accidents. In cases of a mass movement of pedestrians, this prototype notifies the driver to slow down and make way for pedestrians. The vehicle also uses a warning system that helps the user identify some of the busiest streets for pedestrians and cyclists. It will also identify aggressive behavior along roads and highways. For instance, in cases of mass protests the vehicle can have a warning system to warn the user of impending danger. This will incorporate the use of internet software to help the user recognize areas to avoid such as an accident or crime scenes. This pre-warning system will also save the user time from using inaccessible streets. Such software will be updated frequently to promote greater autonomous efficiency. This system will help reduce the need for traffic police to coordinate traffic in busy intersections and urban areas.
The prototype of self-driven cars resonate huge milestones in a technological project undertaken by Google to address the challenge of efficiency in transport system. This model should be geared towards developing autonomy on powered locomotion. The prototype is anchored on the design that powers the car in a manner that it is chauffeur driven (Osipychev et al. 983). Essentially, there is a need for the effective legislation for the sake of approving the public usage of the self-driving cars.
The self-driving experience is referenced on the impact these models will have on the general society. It follows to note with a fleet of 20 vehicles, the prototype of the self-driven, it is clear that people can cover significant milestone of autonomous driving. The system draws its effectiveness from the limited speed that Google stores on the vehicles navigation maps. With regard to this, it is plausible to note that the efficiencies of this prototype stem from the ability often-autonomous cars to maintain sufficient distances between other cars and itself within the spectrum of sensor system (Bimbraw 197). In so doing, the pathways of sensors initiate the possibility of the systems to override with regard to safety. Indeed, this is outside the abilities of the human driver to struggle with controlling the car through steering and braking system or even turning the wheel.
The technology allows people to have access to the use of autonomous driving to a greater sense. Unlike the traditional systems of transport, the prototype of self-driving, the issues of privacy, convenience and accessibility are achieved in the framework of the driverless car fitting in the mental and physical state of the users. In light of this, it is plausible to note that such a system of the care will guide the paradigm in which people without the ability to drive traditional models of car can access and use the driverless cars. Although the burden is on the society, the prototype addresses the primary need of chauffeuring senior citizens, old people, and the sick to go about their duties without fear of convenience (Liu et al. n.d). The programmed routes and their affiliation to gas stations, restaurants, and other social amenities will fit everybody regardless of the underlying social and physical challenges. In fact, this foundational advantage depicts the benefits of technology to the users. In so doing, the challenge of accidents is addressed within the pathways of navigation and programming.
The driverless vehicles come in both partial, while they are completely automated. Therefore, the aspect determines whether they require the intervention of drivers and taking control in the event there are technical and systemic hitches (Bimbraw 198). With the rules of liability and insurance, coupled with the regulations of the government, and other players, the society must appreciate the imperative need to develop the much-needed infrastructure to support the autonomous vehicles. Regardless of such procedures and legal provisions, the prototype fits in the framework of reducing the rate of accidents and the cost of fuel significantly. In future, the prototype of driverless car will allow people to enjoy the safety feature within the prototype of numbing detail. In essence, the detailed sensors can capture up to 200 meters in all directions, thereby eliminating the possibility of blind spots. The prototype allows for defensive and active driving while avoiding other cars, their blind spots, as well as other road users. The flexibility of this prototype enhances the critical systems, avoiding redundancy and providing braking and steering backups.
Despite the challenges and benefits of autonomous vehicles, more companies and governments are considering the future of self-driving cars. Researchers such as MIT are constantly working on technology to improve transport systems for autonomous vehicles. MIT is working on futuristic driverless car intersections (Ulrich). These intersections are designed to reduce traffic reliance on traffic lights and traffic police. Therefore future transport systems will operate on Smart and signal free technology. Future technologies will detect body movements across the roads and other cars. Green technology will also be integrated into driverless car systems. Due to greater environmental awareness, sustainable automobiles will also use solar technology to reduce environmental pollution. More internet based software will also be incorporated into the central systems and engineering of autonomous cars. Manufacturers will also focus on increasing mileage by integrating new smart car technology and car body systems. Higher technological efficiency in autonomous vehicles will help reduce road accidents and promote transport efficiency worldwide.
Self-driving cars continue to change the way people live with technology, promising a greater level of efficiency, safety, and fast travel. As a cutting age of machine learning, robotics, and computer vision, the technology maps innovative ways to integrate automotive regulation. The surge in companies that obtain autonomous testing shapes the faster growth in the number of vehicles that are driverless. The new prototype will also detect pedestrians and cyclists on the road as they cross. The sensor technology is also designed to detect traffic lights across urban areas that indicate pedestrian crossing. Federal and state licensing infrastructure also limit the use of self-driving cars. The government has the responsibility to guarantee the safety of citizens. However, the development of autonomous cars creates challenges in defining safety and the most efficient use of autonomous vehicles among the users. The prototype will provide the promised reality in self-driving vehicles that are needed to save the cost involved in road carnage and time wasted in traffic congestion. However, adequate testing is critical to eliminate the limitations in the prevailing systems.