It’s a funky old Friday again, so once more, it’s time for fascinating stuff from the wonderful world of science.

First, we return once more to one of my all time favorite avenues of technological research. Long time readers with particularly keen memories will already know which one I am talking about. It’s been a fascination of mine ever since I was a kid watching the short lived sci fi show Automan.

I am talking, of course, of self driving cars. Woo hoo!

The latest development on that concerns some whiz-bang propellerheads over at MIT who have developed a system that prevents collisions by predicting the movements of other vehicles.

Right now, it’s proof-of-concept at best, but it’s still an impressive achievement. They took two remote controlled cars, put them on a looped “death track” guaranteed to put them in serious collision peril, and ran one hundred trials of their new system, and there was only one collision.

Still one more collision than we would want in a real world application, but not bad for a new kind of system.

What I particularly like about this system is the excellent methodology. Basically, they systematically broke all the aspects of driving down to their most basic elements, modeled those elements, and from that developed a predictive model that generates all the possible positions a vehicle can be in within the next few seconds, given its current position and trajectory.

They even included factors like whether or not the vehicle is in an intersection or on onramp.

I am very impressed that they were able to tackle such an enormous number crunching task and produce not merely useful results, but actually quite good results. Of course, part of the reason these sorts of tasks can succeed when they failed drastically in previous areas is that we now have computer chips capable of doing the staggeringly huge number of calculations required to do such impressive predictive modeling.

But still, hats off to the prime nerds at MIT for making it work!

Next up, a story I will admit is only science adjacent rather than directly science related, but still, I thought it was worth sharing with you nice people. It caught my eye both because it happened in the marvelously laid back city of Portland, Oregon, a place I lived for a while, and because… well, you will see.

It’s a story of ecology, nephrology, public works, and the difference between science and pragmatism.

The short version : drunk guy pees in one of the Portland reservoirs. Solution : flush the entire freaking eight million gallons of perfectly good drinking water at a cost of almost $40K.

Obviously, strictly in terms of actual public health threat, this is a massive overreaction. For one thing, urine is sterile, and even if it wasn’t, one guy’s drunken whizz is not going to be more than, at most, a quarter of a gallon going into 8 million gallons of water, thus making it one 32 millionth pee.

There’s homeopathic remedies that are stronger than that.

And really, do you think that open reservoirs don’t end up with a lot more pee (and worse) than that simply from the local biologically active wildlife? Realistically?

But of course, this goes into the realm of human taboo, and that’s not a reasonable thing. Were I the administrator involved, the most important factor would be, basically, does the public know this guy pissed in the reservoir? Because if they don’t know, and are not likely to know, then to me, the cost of replacing all the water is not justified.

But if they do know or are quite likely to find out, then you have no choice, only the full flush will do. The public has to be absolutely sure their water is clean, not just from a scientific or reasonable point of view, but from the point of view of our potent sense of taboo and disgust.

We’re not robots, after all. We’re irrational humans, and when it comes to those most profound of our taboos, the ones involving bodily wastes, there is almost no room for negotiation.

Flush that thing. And we shall never speak of this again.

Getting back to full on science, we have an exciting development in that rapidly blossoming real actual no longer science fiction field of nanotech : a self powering nanotech machine that can transmit wirelessly.

It derives its power from any source of vibration, which is nothing new. There have been vibration harvesting nanomachines for a few years now…. decades in nanotime.

But that’s all they could do. Keep going. You had to put power into them from the outside in order to even verify they were still working.

What makes this one the new hotness is that it has enough power to transmit a wireless signal all on its own, throwing the door wide open for all kinds of self-powering nanosensors that could send information from anywhere at all, forever.

After all, damn near everything vibrates, as optical astronomers trying to take long exposures will readily tell you. And once you get down to nanoscale, even the silicon molecules in the heart of Everest vibrate.

The applications for such eternal nanosensors are innumerable. A world where such sensors are cheap and plentiful would be an information-dense world, with trillions of these sensors feeding information to whoever wishes to look it up.

It’s this kind of thing that makes me think “I am truly living the future”.

It’s a wonderful feeling.