Disclaimer—This paper partially fulfills a writing requirement for first year (freshman) engineering students at the University of Pittsburgh Swanson School of Engineering. This paper is a student, not a professional, paper. This paper is based on publicly available information and may not provide complete analyses of all relevant data. If this paper is used for any purpose other than these authors’ partial fulfillment of a writing requirement for first year (freshman) engineering students at the University of Pittsburgh Swanson School of Engineering, the user does so at his or her own risk.
RADAR AND LIDAR DETECTION IN AUTONOMOUS VEHICLES Max Pelna, firstname.lastname@example.org, Bursic 2:00, Julian Linkhauer, email@example.com, Mahboobin 4:00
Abstract—Autonomous automobiles are presently on the forefront of innovation in the realm of transportation and civilian way of life, catching the eyes of top companies and researchers. RADAR and LiDAR play vital roles in the operation of a computer-controlled automobile. The following paper will analyze the effectiveness of RADAR and LiDAR as detection systems in autonomous automobiles and delve into the pros and cons of using RADAR and LiDAR as primary means of detection in regard to other candidates. This technology incorporates the vast majority of engineering disciplines. From mechanical engineers that aid in the creation of the physical product, to computer engineers who design the software and computer systems that control the inner operations of the vehicle, all disciplines have their places in the creation of the autonomous automobile. Major companies such as Lexus, Google, Mercedes-Benz, and Audi are presently applying the technologies referenced in this paper to develop a new flavor of automobile. This paper will begin with an overview of the overarching technologies, LiDAR, RADAR, obstacle detection, and autonomous vehicles, will be presented. To follow, an analysis of the effectiveness of radar as a rapid-response detection system will be provided and reinforced by data synthesized by leading corporations and researchers in the field. Some of the ethical dilemmas regarding the system in question will be addressed and analyzed, and a conclusion on the matter will be drawn. This paper will utilize numerous scholarly sources, official databases, and educational references in order to support claims. In addition, this paper will address throughout the issue of sustainability in regard to this technology. Key Words – Autonomous Vehicles, LiDAR, RADAR, Obstacle Avoidance, Sustainability THE ROAD LESS TRAVELED- FOR NOW Since the inception of the mass-produced vehicle, the evolution of cars has been gradual. As the years progressed, safety has always been a major concern for automobiles. Car manufacturers introduced numerous universal safety features, such as seatbelts, airbags, and crumple zones, which are areas of the chassis that deform to absorb the impact of a collision. In the 1990s, select auto manufacturers like BMW, Audi, and Mercedes-Benz began incorporating luxury and cutting-edge technology. These companies did so while of course keeping the safety of all occupants in mind. This shift in the reasons behind the design of the everyday consumer’s car began most notably in the late 1990s Toyota vehicles. In particular, the 1997 Toyota Camry implemented LiDAR, or Light Detection and Ranging, in cruise control functions. More recently, LiDAR and RADAR have been observed in 2009 and newer Ford and General Motors (GM) vehicles with the introduction of backup cameras (which were mandated in 2014 and will be required in all 2018 model year vehicles), built-in GPS and computer systems, blind spot monitors and more .
These innovations take us to where we are now. A “smart” car, or a car that can communicate and act in response to specific scenarios when necessary is the next technological stepping stone to a vehicle that can operate sufficiently without the aid of a driver. This is an unprecedented undertaking as driverless cars were viewed as science fiction a mere twenty years ago. Autonomous cars under current development are challenging the very nature of automotive safety and innovation, as these cars must meet the expectations of a human operator. One of the most basic necessities of a vehicle operator is the ability to recognize obstacles and act accordingly while also avoiding complications that arise. Therefore, an obstacle detection system that exceeds standards of safety and reliability is imperative. Leading designers are experimenting with a multitude of varying obstacle detection systems in an effort to determine what system is both reliable enough and safe enough to be accepted by the public and safety officials.
Sustainability can be interpreted as the environmental and economic impact of a technology on the world. By this definition, obstacle detection systems that have the potential to make autonomous cars a reality must be able to last. New technology can only be a definitive answer if manufacturers can continue to make it without significant consequences on the environment or consumer economy. If it is not sustainable, it will not take hold. RADAR and LiDAR based obstacle detection systems have both positive and negative effects on the sustainability of the automobile, both of which could dictate its future. Sustainability is not of major concern for the technology itself, more so for the application of the technologies in the autonomous car.
RADAR and LiDAR are two of the most relevant potential foundations for an effective and sustainable obstacle detection system. Each respective detection system has numerous benefits and drawbacks that designers must take into consideration. A RADAR and LiDAR-based detection system is theoretically, experimentally, and ethically a leading contender for a place in obstacle detection for the autonomous car. This brings up two important questions: What is RADAR? What is LiDAR?
AN OVERVIEW What is RADAR and How Do We Use It? Radio Detection and Ranging, or RADAR, is a means of detection that spots objects by emitting short radio waves in some direction . These waves are emitted from the source, where they travel and reflect off of the surface of objects in their path and are received by the source as they return. This, in turn, provides a series of points mapped out in arrays. Each point represents a location where a radio wave has reflected back to the source. The concept of RADAR as a form of obstacle imaging was most notably adopted into mass implementation throughout the World War II Era. Military forces at the time used RADAR to detect enemy ships, aircraft, and tanks in order to gain a strategic advantage over the opposing forces .
While RADAR is most famously known for its applications in the military, space exploration, and meteorology, recent changes to automotive technology have seen the implementation of RADAR for numerous reasons. For example, the 2012 model of the Audi A8 came equipped with a RADAR-based system called Adaptive Cruise Control (ACC). This system detects the vehicle in front of it and locks onto it, allowing the car to automatically adjust its speed to the traffic ahead.  It was not a prototype, nor was it customized; the system came with the car. Other examples of the implementation of RADAR in modern automobiles include auto-park assist in the 2009 Ford Escape Sport Utility Vehicle (SUV). This vehicle was the first of its kind, as no factory-model vehicles were equipped with the ability to park on its own . In addition, the 2012 Mercedes-Benz M-Class SUV has a similar feature. This particular vehicle, however, stood out, being the first vehicle mass-produced to have a park-assist feature that not only parks the vehicle, but “unparks” the vehicle, pulling out of the space .
Vehicles such as these mark milestones in the innovation of vehicles and RADAR alike. The use of RADAR in park-assist, obstacle detection, and a multitude of other applications, is a stepping stone toward the production of a vehicle that is entirely safe and self-operable.
What is LiDAR? Light Detection and Ranging, or LiDAR, is similar in principle to RADAR. The source emits infrared photons in pulses that, just as radio waves, bounce off of objects and return signals to the source. Unlike RADAR, however, LiDAR maps the returned signals into a three-dimensional plane of the basic surfaces the photons reflected off of .
FIGURE 1 
What LIDAR Sees One of LiDAR’s most relevant uses for non-military and non-government agency purposes is its applications in traffic monitoring . Police speed guns that are used primarily in high-speed situations, such as major freeways, are LiDAR-based, as a large amount of emitted photons can be processed and utilized to determine the speed of a vehicle . LiDAR’s accuracy and speed allow it to function very effectively and without hesitation, which are qualities required for rapid maneuvers. From the Present to the Future We have detailed the concepts. We have shown how automotive manufacturers use them to date. Now, we must consider how RADAR and LiDAR can be utilized in the autonomous vehicle. Autonomous cars must be both safe and efficient. Conquering such ethical and technological barriers requires taking into consideration the utmost standards of safety and technology. The aim of the autonomous vehicle is to make the journey from Point A to Point B safer and with improved quality of life. In order to allow for this to be possible, the vehicle must be able to quickly recognize and react to obstacles, threats, and other scenarios that arise during any given journey. RADAR and LiDAR have the potential to allow for this, both independently and as a combined system. Now, we must consider how we can implement RADAR and/or LiDAR in the defense mechanisms and operations of the driverless car to ensure the most efficient and safe trip?
DETECTING THE OBSTACLES TO SAFER DRIVING Drive Safely “Drive safely” is an imperative given by Garmin, Waze, and every other GPS system that provides routes and directions. A major concern for automobiles since their creation has been safety. The automotive industry has already seen the introduction of airbags, seat belts, lights, reflectors, mirrors, and other safety features, become common in vehicles. With modern technology, the possibilities for improvement are endless. Designers are beginning to include fisheye back-up cameras (as per the mandate previously mentioned), real-time assist such as OnStar, side-impact airbags, anti-lock braking (ABS) systems and, as mentioned before, crumple zones, in order to improve the overall safety of automobiles. Innovations such as these provide numerous countermeasures to allow the operator to either avoid a collision entirely or collide with a greatly reduced risk of injury.
However, even with a multitude of safety measures standard in most old vehicles (mid-1990s and earlier model year) and all newer vehicles (late-1990s and newer model year), 2014 saw approximately 30,000 fatal motor vehicle accidents (MVAs), resulting in a number of deaths exceeding 32,000 in the United States alone . While these numbers have decreased in terms of MVA fatalities per capita over the last century, there is still a long way to go. With the introduction of RADAR and LiDAR-based safety measures over the last decade, the number of deaths in car accidents has decreased significantly. In 2004, before RADAR and LiDAR-based systems were commonplace in automobiles, there was an estimated 42,000 fatalities as a result of MVAs . While a direct correlation between the implementation of RADAR and LiDAR-based systems and the reduction of MVA deaths cannot be concretely made due to the vast number of variables, any increase in safety, we can assume, can reduce the probability of collision and is therefore beneficial to driver safety.
Sustainability ties directly into consumer safety. Fewer MVAs means fewer damaged products. When a car is totaled in an accident, it goes to a scrapyard. At this point, all the vehicle does is take up space. While these vehicles occasionally produce scrap that can be used in other vehicles, the amount of damaged parts vastly outweighs the amount of reused parts. Even when replacement parts are in demand, more often than not the parts come from manufacturers that produce the replacement part, not from salvaged vehicles. If autonomous vehicles using RADAR and LiDAR based detection systems are mass produced, the aforementioned statistics would remain true. This means that less vehicles will be crashing, and those that do end up in collisions will have less damage overall than consumer-operated vehicles. Any system that reduces the amount of waste produced by car crashes is beneficial for both the environment and the consumer’s wallet.
Making the Autonomous Car Worthwhile In order for the autonomous car to succeed in mainstream society, it has to be worthwhile for the consumer. If the everyday consumer could purchase a much cheaper car that he can operate by himself, why should he go the extra mile and dish out thousands of dollars more in a car that drives on its own but offers no real benefit? The answer is simple; he shouldn’t. In order for the dream to become reality, the car itself to be our chauffer, the automotive industry to really drive it home, designers must be able to make it worth the extra green from your pocket. Car companies can do so by ensuring a leisurely drive with the utmost safety, far safer than any human could make possible. Detection systems are integral parts of a safe and sustainable autonomous car. RADAR and LiDAR-based systems can pave the road to success. DETECTION: SEEING THE FUTURE RADAR in Autonomous Vehicles As previously stated, RADAR uses radio waves to detect both stationary objects and objects in motion. Autonomous cars must come equipped with a detection system similar to the Audi A8’s adaptive cruise-control system which, as stated, uses RADAR to detect vehicles further down the road and adapt vehicle speed so as to follow safely . Such a system must be applied to a greater extent in the autonomous car. An obstacle detection system configured to use radio waves emitted in a full panoramic view would allow a similar result, accounting for all 360° around the vehicle, as opposed to directly in front. This could be done by coding a program into the car’s computer system, similar to that which dictates how the Audi reads and responds to vehicles ahead of it, but amplified to function in all directions. The computer would take the data input from the sensors’ map of perceived objects, process their distance relative to the exterior of the car, and react accordingly.
For example, this technology can be seen on a smaller scale in the 2013-16 model year Mazda 6 sedans. This particular car is fitted with blind spot monitors that use RADAR . The car sends out radio waves, in directions predetermined by the manufacturer, which scan the blind spots specific to that vehicle. If another car is detected within 20 feet of the car, a sensor will go off that, in turn, illuminates an icon on either the left or right wing mirror, depending on which side the obstacle is present, which informs the operator of an object in his or her blind spot as seen in figure 2.
FIGURE 2 
Blind Spot Monitor as Seen in the 2016 Mazda 6 Sports Sedan This same technology can be applied to autonomous cars by scanning in all directions with an amplified range via an external antenna similar to those seen on older vehicles as seen in figure 3.
FIGURE 3 
Modern Car Antenna as Seen on 2009 Chevy Captiva and 2008 Saturn Vue Antennae can increase the car’s ability to receive radio waves it transmits, just like they are used to receive signals from radio stations. These antennae can be used to receive radio waves bouncing off of more distant sources, allowing better obstacle detection at both higher speeds and greater distances. If the autonomous vehicle senses traffic rapidly slowing down, for example, it needs adequate time to react. RADAR, amplified to scan a significant distance around the vehicle, can provide this necessary amount of time. Radio waves travel the speed of light, or 186,000 miles per second, and can therefore provide a nearly instantaneous reaction to obstacles it detects and processes.
Pros and Cons of RADAR RADAR emits radio waves at the speed of light. Since it functions at such a high speed, it is quite simple for RADAR to provide stimuli to the car’s detection system rapidly. This allows the car to act quickly should an event arise where fractions of a second could mean the difference between avoidance and collision. In addition, RADAR systems are not all that complex, as a basic configuration includes just a transmitter, receptor, and computer for processing the data. RADAR is statistically resistant to low visibility, such as darkness or inclement weather. Because radio waves have a much larger wavelength than visible light, they are less effected by obstacles that would generally hinder light from passing as seen in figure 4 .
FIGURE 4 
A Radio Frequency Energy Wave Superimposed upon an Infrared Energy Wave
In contrast, RADAR is more primitive in the sense that it cannot display what it sees accurately, which is made evident by figure 5 on the following page..
FIGURE 5 
Typical RADAR Display RADAR, by its very nature, cannot display all visual aspects of the objects it detects . Consider RADAR and LiDAR detecting a chair. RADAR can accurately pinpoint where the chair is located but when one looks at its display, it shows a vague shape depending on the complexity of the display. LiDAR, on the other hand, will show both location relative to the transmitter and a more specific shape of the object as previously seen in Figure 1. In addition, RADAR antennae pick up interference from other sources, such as broadcasting stations, outer space, and other RADAR sources. However, this can be minimized by ensuring the wires contained in the antenna are properly shielded and the source of interference is not incredibly close and strong. In other words, unless you plan to drive into a radio station, shielded wires should be sufficient.
RADAR also has a significant disadvantage in navigation, as it cannot read road markings that are flush with the pavement and not distinguishable from the surface of the road. RADAR can theoretically detect road markings, provided one of two requirements are met. Either the road must have elevated markings that can be detected by a RADAR-based system, or the system must be able to detect minute distances, such as the small height differences painted markings provide. That level of sensitivity, however, complicates the system even further, as it may detect small impurities in the road or small objects such as pebbles and attempt to react to those unnecessarily.
LiDAR in Autonomous Vehicles LiDAR, generally, is newer to the automobile than RADAR. LiDAR, unlike RADAR, uses cameras and lasers that eject photons but otherwise functions similarly to RADAR. LiDAR sends out photons in the form of lasers that reflect off of surfaces and, in turn, map out an image of what the photons detect. The internal computer of the vehicle takes these images (see figure 1) and scales them, applying relative distances .
Recently, LiDAR has taken on new uses in the automobile. While luxury cars with adaptive cruise control generally use RADAR, more budget-friendly vehicles tend to utilize LiDAR. For example, the 2016 Chrysler 200C uses a small camera nearly identical to a back-up camera . In this case, the camera scans upcoming obstacles, transmits the data to a computer built into the car, and takes corrective action if needed . For instance, if the owner of the car is driving down the highway and traffic suddenly slows because someone without blind spot monitors caused an accident, the camera can recognize this and communicate with the car. The car, in turn, will apply the brakes if necessary to help avoid collision as seen in figure 6.
FIGURE 6 
Adaptive Cruise Control Display as Seen in Audi A8
Autonomous cars can utilize a LiDAR-based obstacle detection system in ways similar to that of a RADAR-based system. While a single camera lens can only capture images of a specific region, a spherical lens can capture images in all directions. This type of lens could be created by containing numerous small cameras in a semispherical shape, similar to that pictured in figure 7.
FIGURE 7 
Bubl Brand 360-Degree Camera Optimally, a lens would be mounted on the roof or some elevated flat surface of the car so as to scan in a full circle around the vehicle, as well as at a slight downward angle . This downward angle is important, as it can see objects that are closer to the exterior of the car. The images produced by the cameras would be processed by the autonomous car’s internal computer. Much like a RADAR-based system, designers would code a program that can rapidly differentiate what is a threat and what is not a threat and respond if necessary. This program would include parameters such as a minimum distance from the vehicle, direction of travel, and size of obstacle. If the program determines that a threat is imminent, it can tell the vehicle to react by a variety of means, including steering, braking, or accelerating. What classifies as an imminent threat could include an obstacle in motion interfering with vehicle trajectory, a stationary object within the vehicle’s path, potentially hazardous debris in the roadway, or a great risk of collision from an object approaching the vehicle.
Pros and Cons of LiDAR LiDAR, being based off of photons, travels at the speed of light. Just like RADAR, LiDAR’s high speed capabilities allow for rapid reaction to threats and situational changes. LiDAR, by nature, uses photons that generate images so the computer can process what is captured . This allows a much more advanced display than RADAR, as it shows a three-dimensional plane. That chair (mentioned previously) has shape with LiDAR as opposed to a singular point with RADAR. In addition, the LiDAR system can come with a database of predetermined obstacles, such as road lines, curbs, vehicles, and pedestrians, similar to how a digital camera can recognize individual faces in an image .
LiDAR uses photons of infrared light, which have a significantly lower wavelength than RADAR, as seen in figure 4 . Because of this, the signal is much more prone to interference from reduced visibility situations, such as inclement weather. In addition, LiDAR also is affected by substandard road markings, as it cannot distinguish markings that are too faded and do not stand out significantly from the rest of the pavement. This consequence, however, is much less drastic than the complications that arise from RADAR-based detection in the same situation.
RADAR AND LIDAR, HAND IN HAND RADAR and LiDAR: The Dynamic Duo? When compared, RADAR and LiDAR appear to overlap. Where RADAR falls short, LiDAR could pick up the slack, and vice versa. To support this statement, consider the following scenarios: inclement weather and city driving. In inclement weather, an autonomous vehicle powered by a LiDAR-based detection system has a significantly reduced visibility and the occupant dies when his car plows full-force into the back of a jack-knifed truck that was obscured by the snow. However, another autonomous vehicle, powered by a RADAR-based detection system slows significantly and pulls onto the shoulder, as it detected the same semi through the snow that shielded LiDAR’s vision. In another example, a RADAR-based autonomous vehicle is navigating through downtown New York City and causes gridlock when it turns from the straight lane and collides with a taxi on Broadway. At the same time, on Wall Street, a driver is enjoying the sights as his car carefully navigates the city streets on its own. In each situation, one of the two forms of obstacle detection has a clear advantage over the other. This raises the question; can we combine the two?
Comparing and Contrasting RADAR and LiDAR Combining RADAR and LiDAR into one system counteracts numerous flaws in each respective technology. Where RADAR lacks the ability to aid in navigation of the roadway, LiDAR comes through with flying colors. Similarly, where LiDAR lacks the ability to detect obstacles effectively in low-visibility scenarios, RADAR picks up the slack. Implementing both methods of obstacle detection into one system allows the system to pick and choose which system is most beneficial, based on parameters designers can code into the program. For example, designers can have the car navigate using only LiDAR until visibility drops below some predetermined, arbitrary distance in front of the vehicle, at which point RADAR can kick in and take over long-range detection. Of course, LiDAR would have to continue operating in order to distinguish road markings, which RADAR has little to no capability of doing. In addition, if one system’s performance is substandard, the other system can take over the workload until repairs can be made.
One major drawback to this system lies in the processor. While both systems function at approximately the speed of light, the computer must be able to rapidly process the data it receives and react accordingly, all in less time than it would take the average driver. If the computer is not fast enough, it will not matter how fast RADAR and LiDAR work. In addition to this, neither system is a complete failsafe for the other. Therefore, if one system fails, some operations and capabilities would be halted. This could be solved by implementing an emergency backup system that allows the occupants to gain control of the vehicle and operate it safely, should the obstacle detection system fail.
While such a system or combination of systems could allow autonomous cars to become reality, we must consider numerous factors. Do the pros outweigh the cons? Can a car driven by a computer really be safer than a car driven by the trained human operator? Is leaving vehicle behavior in the hands of a computer ethical?
EFFECTS ON THE ENVIRONMENT AND ECONOMY Sustainability is of major importance for any significant technological advancement. If it is not effective both environmentally and economically, it no longer is a step forward. Autonomous vehicles running with RADAR and/or LiDAR based systems are statistically safer than a human-controlled vehicle. This results in multiple benefits and drawbacks to sustainability. MVAs There are numerous consequences to a reduced number of MVAs. Firstly, the environment benefits from a reduction in the number of accidents because fewer collisions results in less waste. If crashes are less severe and in fewer numbers, there will be fewer totaled vehicles and damaged vehicle parts. Damaged parts and totaled vehicles go directly to scrapyards and landfills, effectively dooming thousands upon thousands of tons of steel to sit there, unused. The majority of parts used in vehicle repair come from the manufacturer directly. Manufacturers produce replacement parts to ensure the greatest quality repair job, as scrap parts are generally do not meet quality standards. Autonomous cars that can effectively avoid crashes therefore reduce the amount of waste produced by MVAs, reducing the amount of material wasted and therefore helping to preserve the environment. While the amount of waste reduces, the amount of money in the average consumer’s bank account increases. When there is less damage to vehicles and less need for repairs and replacements, the consumer saves money.
While the environment benefits from a reduction in the number of damaging crashes, some people suffer economically. Insurance companies that generate the majority of their revenue from increased rates after automobile accidents now lose money. In addition, car dealerships, mechanics, and auto-body specialists lose money and business as there are less cars that require service. Infrastructure The most significant obstacle to the sustainability of the autonomous car is adaptation of infrastructure. Autonomous cars will require adaptations to infrastructure. Using a RADAR and LiDAR-based obstacle detection system requires easily distinguishable road markings, such as raised, illuminated, or recessed markings. This is the case because the detection system cannot easily detect a marking flush with asphalt. Adapting our current roadway infrastructure would be a costly but essential investment, as it would require marking every marked road with lines and markers that a RADAR and LiDAR based system can detect. Once the infrastructure is in place, however, the costs are comparable to the present day road maintenance costs. THE ETHICS BEHIND THE WHEEL It is a general trend that automation is innovation. The less human intervention, the better. However, are we right to assume that to be the case with autonomous vehicles? Ethics in the context of autonomous vehicles are the requirements that the vehicle meets safety, efficiency, and sustainability standards. These are the overlying ethical dilemmas surrounding the application of autonomous vehicles.
Fail-Safes One day, a man is driving to work in his new autonomous Ford Bronco when he notices the vehicle beginning to slow down and drift out of the passing lane. On the small dashboard, a camera icon illuminates, informing the passenger that the obstacle detection system keeping his car on a safe trajectory has failed. Because there are no fail-safes correcting this issue in an emergency scenario, that poor man’s Bronco ends up crossing the median and colliding with an oncoming car.
A scenario such as this proves that, ethically, the system alone is not enough. In order to bypass this ethical dilemma, designers must incorporate numerous safeguards. For example, the vehicle could theoretically come equipped with a small remote control, consisting of a joystick that rotates a full 360°. This controller would act as a means of controlling the car in case of emergency. Revisiting the scenario with such a device installed in the car, the Bronco owner, after noticing his car slowing down and drifting, grabs the remote control from the center console and guides the car to the shoulder of the highway, avoiding catastrophe and traffic issues. The same scenario could be revisited in a similar manner using a built-in steering column that is inactive until the detection system goes offline. Once this system is no longer active, the vehicle could be programmed to enter a manual mode, where the operator has control of the vehicle and can maneuver it safely to a stop.
Combination Systems LiDAR and RADAR-based systems both have significant flaws. As previously explained, these flaws overlap, meaning that where LiDAR falls short, RADAR can keep running and vice versa. So, is a combination of both a LiDAR system and a RADAR system not the most ethical approach?
For example, car A has an obstacle detection system based on LiDAR. Car B has an obstacle detection system based on RADAR. Car C has both systems combined and working in unison. Car A is driving down a street on a foggy morning. Car B, having no means of detecting road lines, crosses the double yellow and is travelling directly into the path of car A. Car A does not recognize the situation because the LiDAR system is incapacitated due to drastically reduced visibility. Car B then collides with car A. Car C, following closely behind car A, detects a sudden decrease in car A’s speed, due to RADAR’s increased immunity to inclement weather, and quickly detects the line indicating the shoulder directly to the right of the car, thanks to LiDAR’s ability to detect flat road markings via camera, similar to face recognition on a digital camera. Therefore, car C quickly swerves onto the shoulder, avoiding collision with cars A and B. This scenario suggests that the only ethical option is a combination of both RADAR and LiDAR in obstacle detection.
RADAR AND LIDAR- THE FAST LANE TO SUCCESS RADAR and LiDAR both have their ups and downs. RADAR uses radio waves to detect obstacles and functions much better in lower visibility scenarios than LiDAR, but cannot distinguish exact structures or road markings. LiDAR, on the other hand, uses lasers and cameras to detect obstacles and can see road markings, but has substandard sight in low-visibility situations. Combined, the two balance out numerous flaws each respective system presents and is, both statistically and theoretically, the most ethical option. Implemented correctly, with all corrective measures taken into account and fail-safes put to the test, a RADAR and LiDAR-based combination system is the one that will both make it and brake it.
REFERENCES  G. Wallace. (2014). “U.S. requires new cars to have backup cameras.” CNN Money. (online article). http://money.cnn.com/2014/03/31/autos/rear-facing-cameras/.
 S. Foley. (2011). “World War II Technology that Changed Warfare – Radar and Bombsights.” Academic Symposium of Undergraduate Scholarship. (online presentation). http://scholarsarchive.jwu.edu/ac_symposium/8.
 R. Stevenson. (2011). “Long-Distance Car Radar.” IEEE Spectrum. (online article).
 L. Solomon. (2006). “LIDAR: The Speed Enforcement Weapon of Choice.” Law Enforcement Technology. (online article). http://webcache.googleusercontent.com/search?q=cache:gFRaeNW-y5UJ:www.officer.com/article/10250592/lidar-the-speed-enforcement-weapon-of-choice+&cd=4&hl=en&ct=clnk&gl=us
 (2014). “General statistics.” Insurance Institute for Highway Safety and Highway Loss Data Institute. (online article). www.iihs.org/iihs/topics/t/general-statistics/fatalityfacts/overview-of-fatality-facts.
 (2004). “Traffic Safety Facts 2004.” U.S. Department of Transportation. (online article). http://www-nrd.nhtsa.dot.gov/Pubs/TSF2004.pdf.
 “How to: Use Park Assist (ML 2012).” (2011). Silver Star Mercedes- Benz. (video). https://www.youtube.com/watch?v=jm0NeO7awkA.
 K. Williams. (2015). “Road Tripping in the 2016 Mazda6 Sports Sedan.” The Vacation Gals. (online blog). http://thevacationgals.com/road-tripping-in-the-2016-mazda6-sports-sedan/.
 M. Pelna. (2015). “2009 Chevy Captiva.” (picture).
 M. Kaine-Krolak, M. E. Novak. (1995). “An Introduction to Infrared Technology: Applications in the Home, Classroom, Workplace, and Beyond …” Trace R&D Center, University of Wisconsin. (online article). http://trace.wisc.edu/docs/ir_intro/ir_intro.htm.
 “Specification of 3D Laser Radar.” IHI Corporation. (online article). https://www.ihi.co.jp/3DLaserRadar/en/product_01.html.
 C. Metz. (2015). “Laser Breakthrough Could Speed the Rise of Self-Driving Cars.” Wired. (online article).
 A. Gold. (2015). “Adaptive Cruise Control.” About Autos. (online article). http://cars.about.com/od/thingsyouneedtoknow/fl/Adaptive-Cruise-Control.htm.
 J. Fingas. (2013). “Bubl’s 360-degree camera records Street View-like spherical footage (video).” Engadget. (online article). http://www.engadget.com/2013/11/05/bublcam-360-degree-camera/.
 A. Iliaifar. (2013). “LIDAR, Lasers, and Logic: Anatomy of an Autonomous Vehicle.” Digital Trends. (online article). http://www.digitaltrends.com/cars/lidar-lasers-and-beefed-up-computers-the-intricate-anatomy-of-an-autonomous-vehicle/.
 (2012). “A Comparison of Lidar and Camera-Based Late Detection Systems.” GPS World. (online article). http://gpsworld.com/a-comparison-of-lidar-and-camera-based-lane-detection-systems/.
ADDITIONAL SOURCES M. Ansari, G. Pannu, P. Gupta. (2015). “Design and Implementation of Autonomous Car using Raspberry Pi.” International Journal of Computer Applications. (online article). http://research.ijcaonline.org/volume113/number9/pxc3901789.pdf.
A. Davies. (2016). “Google’s Self-Driving Car Caused Its First Crash.” Wired. (online article). http://www.wired.com/2016/02/googles-self-driving-car-may-caused-first-crash/.
M. Green. (2013). “Driver Reaction Time.” Human Factors. (online article). http://www.visualexpert.com/Resources/reactiontime.html.
R. Hagemann, A. Thierer. (2014). “Removing Roadblocks to Intelligent Vehicles and Driverless Cars.” Mercatus Working Paper. (online article). http://mercatus.org/sites/default/files/Thierer-Intelligent-Vehicles.pdf.
R. Hudda, C. Kelly, G. Long, et al. (2013). “Self Driving Cars.” Fung Institute for Engineering Leadership. (online article). https://ikhlaqsidhu.files.wordpress.com/2013/06/self_driving_cars.pdf.
A. Kane, P. Koopman. (2013). “Ride-through for Autonomous Vehicles.” Carnegie Mellon University. (online article). http://users.ece.cmu.edu/~koopman/pubs/kane13_ridethrough.pdf.
A. Peseri. (2013). “The Google Driverless Car: ‘A Cool Thing that Matters.’” Humanity Centered Robotics Initiative. (online article). http://hcri.brown.edu/2013/08/25/the-google-driverless-car-a-cool-thing-that-matters/.
R. H. Rasshofer, M. Spies, H. Spies. (2011). “Influences of weather phenomena on automotive laser radar systems.” Advances in Radio Science. (online article). http://www.adv-radio-sci.net/9/49/2011/ars-9-49-2011.pdf.
S. Rathod. (2013). “An autonomous driverless car: an idea to overcome the urban road challenges.” Journal of Information Engineering and Applications. (online article). ISSN: 2225-0506.
(2014). “The road to self-driving cars; Today’s crash-avoidance systems are the mile markers to tomorrow’s autonomous vehicles.” Consumer Reports. (online article). http://www.lexisnexis.com.pitt.idm.oclc.org/hottopics/lnacademic/?shr=t&csi=7944&sr=HEADLINE(%22The%20road%20to%20self-driving%20cars.%22)%20and%20date%20is%202014.
R. Stevenson. (2011). “Long-Distance Car Radar.” IEEE Spectrum. (online article).
ACKNOWLEDGMENTS We would like to thank our chair, Frank Kremm, and our co-chair, Kyler Madara, for helping us improve our paper with constructive feedback. We also thank the writing instructor for our class, Emelyn Fuhrman, who gave a detailed overview of the assignment and the writing instructor that has been grading our assignments, Joshua Lapekas.
University of Pittsburgh, Swanson School of Engineering