I don't think it's especially helpful to use this kind of "let's look at an infinite timeline/every possible outcome" type of reasoning. What if a region's local economy crashes and there are no more cars or pedestrians? Those bollards sure seem like a waste of money now! What if? What if?
There's no algorithm to make this decision. It is best to do it iteratively, intelligently, and wisely. You use a bit of science and statistics, read the room to make a vibes-based analysis of what people want, present the public with a proposal that matches their principles with your own principles, and to finally look at the results after some time. You mention yourself what is basically the 'optimal engineering outcome' is apparently to eliminate pedestrians altogether. If that's what engineering wants, then engineering is wrong.
> I don't think it's especially helpful to use this kind of "let's look at an infinite timeline/every possible outcome" type of reasoning. What if a region's local economy crashes and there are no more cars or pedestrians? Those bollards sure seem like a waste of money now! What if? What if?
That would be pretty stupid. We should do what I'm doing instead and focusing on what the acceptable rate of accidents is.
If you check you'll notice I've talked about literally 0 "what ifs" or hypothetical situations beyond picking an arbitrary 80% to showcase that even an arbitrarily good improvement won't make a difference to the process of demanding improvement [0].
But if we're going to demand 0 deaths then the obvious solution is to completely isolate cars and pedestrians. I've seen some similar work to this in the past, and a lot of money thrown at it didn't come up with a better solution for perfect safety. Even isolation won't actually achieve a perfect rate, but it gets it low enough that people won't cotton on for a few decades and that has to do.
If we're not demanding 0 deaths, then the process we use can't be "see a death -> demand improvements". That is like a while loop with no stopping condition - and we're implicitly heading to 0 with it anyway so we may as well short circuit the needless deaths along the way. Someone needs to explain what the target rate is, or how to figure it out.
> It is best to do it iteratively, intelligently, and wisely.
This is road engineering. Engineers have been building roads for 3 millennia, roads for motor vehicles for more than a century and modern statistics has been settled enough for this sort of work for about the same length of time.
We're way past the point where we need to iterate. Say what is acceptable and the engineers will build it. That, in fact, is likely what happened to get the current rate of deaths and injury - someone did a cost benefit and tried to set design standards as close as possible to an optimum point.
If you want them to optimise for something, tell them what you want and they can build it. There is no need to play games with the civil engineers, it is cheaper to just be upfront with design constraints.
[0] If anything, it makes than backlash worse. I've seen people demand Boeing get nationalised for a safety record that is still better than what happens on public roads. The consensus position seems to be investigations and punishments for Boeing management. There is no pleasing some people.
There's no algorithm to make this decision. It is best to do it iteratively, intelligently, and wisely. You use a bit of science and statistics, read the room to make a vibes-based analysis of what people want, present the public with a proposal that matches their principles with your own principles, and to finally look at the results after some time. You mention yourself what is basically the 'optimal engineering outcome' is apparently to eliminate pedestrians altogether. If that's what engineering wants, then engineering is wrong.