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.
Expert journalist Mark Hoag states that safety on the roads has become a serious priority for car companies, and even tech companies . According to the USA Causes of Death report road traffic accidents are the number one cause of death for fifteen to nineteen year olds in the United States , and over ninety percent of these crashes are due to human error (this includes drunk driving) as per the Stanford Law reveiw. If the human element were removed from the equation, logically, the rate of fatal car crashes would be reduced by over ninety percent as well. Every community has a story like mine, and that should not be true. Autonomous vehicles (AVs) can help prevent this pain, they could mean so much to so many people.
As if that is not a reason enough for autonomous vehicles to be adopted, they have countless other benefits that range from increased efficiency, to reduced need for parking spaces and garages in cities, to decreased green house gas emissions . According to the Global Autonomous Vehicle Partnership website if ten percent of the cars on the road were autonomous more than thirty seven billion dollars could be saved (through less wasted time and fuel) as well as a lower death toll, if ninety percent of cars on the road were autonomous the savings would be around 475 billion dollars .
Autonomous vehicles are coming, they are inevitable, the only question left is when, and how they are coming. In this vein there are two distinctive approaches, that of Google and that of Tesla. One of the biggest differences between Google and Tesla cars is the usage of LiDAR, short for Light Detection and Ranging . Google uses LiDAR in its vehicles whereas Tesla only utilizes cameras and radar as its “eyes”. The development of this technology has led to several innovations with multiple other applications to engineering and the world. This includes Gallium Nitrate transistors, which can be applied across countless other types of technology. AVs, and specifically LiDAR will help save lives, encourage the creation of new technologies, and make sure that there aren’t any more stories like Alex and Calvin’s.
LiDAR is a remote sensing method that uses pulsed laser beams to judge the ranges (distances) between objects . It measures the amount of time it takes for the impulse to be directed back and from that data and integrated GPS, Inertial Measurement Unit systems, scan angles, and calibration it is able to create a bunch of points . It uses that data to create a high-resolution map of its surroundings; LiDAR can normally map its surroundings for one hundred meters each way .
LiDAR can have two types: topographic and bathymetric . Topographic LiDAR utilizes light at a wavelength close to that of infrared light and is used to topographically scan the earth . According to the National Ocean Service bathylimnetic LiDAR is used in the oceans to record water depth . Bathylimnetic LiDAR uses a much higher frequency (shorter wavelength) type of light, a green laser . LiDAR has a variety of applications including the creation of topographical maps One of the most innovative and relevant applications of LiDAR currently is in AVs. In most cars equipped with LiDAR it is placed on top of the car, and rotates very quickly (exact specs unknown but somewhere around 600rpm) . This rotation allows the car to sense its quickly changing environment in high resolution and make decisions . LiDAR technology is used in virtually all AVs being tested currently, excluding Tesla . Despite its commonality in AVs LiDAR comes with its own set of problems .
LiDAR is extremely expensive, currently about eighty thousand dollars for one sensor according to David Krambeck’s research on AV pricing . That is significantly more than the cost of the radar and camera that is used by Tesla AVs . However currently there is a company called Efficient Power Conversion (EPC) which sells Gallium Nitride transistors for use in LiDAR technology (among other things) which would make the LiDAR technology not only cheaper but faster and more efficient . A Startup called Quanergy Systems that is utilizing the GaN transistors is building a LiDAR system that would only cost 250 dollars . Furthermore these GaN transistors have come to the market (and people’s attention) almost completely because of these LiDAR problems , even though at this point GaN transistor purchases for LiDAR manufacturing only account for fifteen to twenty percent of EPC’s sales . These transistors are capable of working at anywhere between one hundred and a thousand times faster than their silicon counterparts . LiDAR is not only an important part of AVs but it is also encouraging engineers to look for solutions to problems that previously they couldn’t because it wasn’t profitable, or because they hadn’t even thought of that problem yet.
Another current negative cited by detractors of LiDAR is that it does not behave the same all of the time. LiDAR’s inaccuracy stems from the fundamental nature of the interaction between lasers and water. Lasers can often times be refracted through water, whether it is in its solid, liquid, or gaseous state . This refraction causes the LiDAR to be unable to correctly detect the distance between objects thus much less reliable in heavy rain, snow, or fog; basically any time there is a lot of water in the air . This seems to pose a huge problem, if LiDAR is only extremely accurate when there isn’t water in the air, what will AVs in Seattle do? But, most AVs with LiDAR are also equipped with However Tesla’s cameras and radar have proven to be even less accurate in practice, just a couple of months ago the first fatality from AVs was recorded when Tesla sensors misread a white eighteen-wheeler and a bright sky [8,9]. It is speculated that a car equipped with LiDAR would have been able to avoid this crash . Each technology has its own limitations hence the need for LiDAR, cameras, and radar. Proponents of LiDAR are not against cameras and radar; they just believe LiDAR is also necessary for a fully autonomous vehicle to be reliable in all conditions.
The final and seemingly largest problem LiDAR creates is the amount of data collected by LiDAR systems. LiDAR sensors collect an obscenely large amount of data, according to wired a square mile LiDAR map can use up several gigabytes of data . This could cause a huge problem in the future seeing as all AVs would be collecting and updating this huge amount of data and a huge storage problem would occur. Fortunately a start up called Civil Maps has created software that is able to turn a terabyte of LiDAR mapping data into only the essential eight megabytes that the car need to function . Google, Uber, and Here are all working on similar software, meaning that soon enough this problem will be negated with their advances .
The Google car is the best example of LiDAR in the context of an actual functioning vehicle. Google uses a Velodyne 64-beam laser . This laser can rotate 360-degrees horizontally  and according to the Velodyne website vertically 28-degrees . The unit has the ability to scan 2.2 million points per second from 120m away with the accuracy of within two centimeters . The LiDAR is mounted on top of the car so as to avoid obstructions and is used to gather these highly accurate readings .
One of the hardest things I used to have to do is drive by a memorial for two high school kids that died in a car crash on my way home from school. The pain so many people feel everyday with the loss of their children, their parents, their friends is unbearable and unacceptable. Autonomous Vehicles are needed to end the unnecessarily high number of traffic fatalities. Engineering itself as a discipline also needs the innovations of LiDAR, and even more specifically the Google Car to move forward with the development of these driverless cars. Google is on the cutting edge of AV technology, and moving towards fully autonomy more rapidly than anyone could have imagined . The byproducts of the efforts to work on LiDAR have already been extremely useful, such as the creation of the GaN transistor. LiDAR itself, although imperfect, will help society realize a world where my children will not have to attend the funerals of multiple classmates who died in distracted driving accidents.