The Uber self-driving car crash on March 18th, 2018 has raised the intense discussion in automation. The reputation of the self-driving car takes a hit since then. What do people think about this technology and this accident? I collected Reddit comment data and try data mining and visualization. But first, We need to know what exactly happened in the Uber crash.
How does the accident happen?
According to the NTSB report, the human driver has no fault in this accident. The human driver was monitoring the automation system and the sensor screen. Even though the driver’s private and business phones were in the vehicle, the proof showed that the driver never used them until she called 911 after the accident. The pedestrian wore dark clothes and her headlight could not be seen due to the angle. The sensor had detected the pedestrian but the AI identified that as an unknown object and failed to warn the driver.
(Left) Location of the crash on northbound Mill Avenue, showing the paths of the pedestrian in orange and of the Uber test vehicle in green.
(Right) Postcrash view of the Uber test vehicle, showing damage to the right front side.
View of the self-driving system data playback at about 1.3 seconds before impact, when the system determined an emergency braking maneuver would be needed to mitigate a collision. Yellow bands are shown in meters ahead. Orange lines show the center of mapped travel lanes. The purple shaded area shows the path the vehicle traveled, with the green line showing the center of that path.
source: NTSB report
https://www.ntsb.gov/investigations/AccidentReports/Reports/HWY18MH010-prelim.pdf
source: ABC News https://www.youtube.com/watch?v=8IqpUK5teGM
What do people think?
The self-driving car is a stable topic in Reddit. People talk about it all year. Uber is not the first brand in the self-driving area but it is one of the most popular brands that have been discussed most.
The number of times people mentioned this brand
After the Uber crash happens, people seem to have more passion for discussion. The number of comments about Uber increases dramatically. Even though people have more negative opinions on Uber self-driving cars in March and April when the Uber crash happened, there are more positive opinions on self-driving cars in all. People see this crash as a failed test killing people but they still have the expectation on self-driving cars.
left three: Reddit score scatter in March, April, and all year in 2018-2019
right top: meaning analysis of Reddit comments in 2018-2019 about Uber
right bottom: daily comments and scores about Uber self-driving cars
The visualization of the most popular words when talking about uber reveals several facts. People would like to compare Uber with other companies in this area.
People care about safety, driving experience, and price. The safety here means not only the driver’s safety but also public safety. This shows that the public still hold the concern and worry about the potential damages caused by self-driving cars and there are technology and moral issues that have not been solved. 
More info: https://public.tableau.com/views/uber_15568234610350/1?:embed=y&:display_count=yes
The graphs and charts above show the public opinion on the self-driving system and automation. I have also analyzed the Reddit comments on the reasons for the Uber crash. Actually, there are many reasons for this accident. The braking system is banned unless the self-driving system allows so. The driver is blamed because she doesn’t use her human eye to see the surrounding and fail to do anything before the crash. The footage or the cameras fail to detect the human figure but lidar and radar should do. However, there was no warning from the lidar and radar.
People also criticize uber as an unreliable company since there are already several accidents caused by Uber.
What people blame on Uber crash
The interesting fact here is that there is a paradox between machine and human. People don’t always trust in machines or algorithm and a human being is expected to do the final decision. However, the decision maker may rely on the data and the machine to make the choice “objective” and “comprehensive”. This leads us to more discussion in the concern about self-driving cars.