It's been released for some days already but just in case anyone have not seen it yet, the youtube link is one of many which contains the video footage released by the police from both the dashcam and the internal cam of the Volvo that was used by Uber's self-driving car test program which hit and killed a pedestrian.
Over the days since the release of the video, I have read quite a few articles and there have been many comments made by people on their personal opinions on the incident. Here, I will NOT put out any of my own opinions but instead will just offer a couple of observations to remind readers of the unique/unusual perspectives of this incident.
1. The most common comment I have read has consistently been how dark the road is and how the pedestrian could not be seen until the last minute. Some have even made comments about her crossing illegally. Almost all universally admitted that they themselves would not have stopped in time because visibility was so bad and the pedestrian was only seen right at the last couple of seconds.
Point : The road -is- indeed very dark. However, it is also one way. And there was virtually no traffic ahead. Under the circumstances, many people, certainly car-nuts will be having their -HIGH BEAM- on. A properly aligned high-beam would illuminate way ahead of the car in such a situation, improving forward visibility tremendously. At the minimum, if available (and most car nuts would have them), lower front fog-lights will definitely on which would illuminate the surface of the road for a great distance ahead.
Why the Uber Volvo was still driving with low beams is probably something to be explained by the Uber team themselves but perhaps it might be related to the 2nd point below.
2. Most comments are from the perspective of the human driver, and is based on human sight and reaction times. I.e. people admits because it came so sudden and so quickly, they too wouldn't have been able to avoid hitting the pedestrian.
Point : The thing is the Uber Volvo is not relying solely on visible light for its operation. A check with Uber's website showed plenty of comments about its use of LIDAR as one of many sensors. LIDAR is based on laser in an invisibile spectrum of light. Additionally because the laser is not visible to the human eye, could it likely be angled higher and now so low as a normal car headlight in low-beam mode ? That would have a greater 'front visibility' for the car's driving computer.
Additionally, a car in the Volvo's market segment would usually also be equipped with numerous other latest safety technology, in particular Collision Mitigation technology and many manufacturers are even putting out pedestrian avoidance type technology. As the video commentators noted, one of the biggest issue authorities have is why the Uber Volvo showed no signs of stopping or even slowing down at all. It's as if the self-driving front sensors totally did not detect the pedestrian which of course be a most worrisome thing.
Keeping at least these two points in mind (definitely there are more to consider but we only cover these two here) would likely be useful to understand the angle and logic from where the final judgement from the authorities are made when they become available.
The youtube video below is from ABC (a US TV network). The anchor image above is also taken from the ABC online website.