Math professor's side mirror that eliminates 'blind spot' receives US patent

While I agree that the sites are useful, HN guidelines request that you submit the original article. That reduces the number of duplicate submissions.

True, but as you get used to seeing the vehicle cross from your center mirror to your side mirror, you get a sense of where it is. Your point of reference just changes from your car in the side mirror to the combination of the location of the vehicle in the side mirror and the center mirror. You really just have to get used to it. I swear by this technique now.

This doesn't replace shoulder checks, so as long as you keep doing that, you should get used to it with a couple hours of drive time.

Just lean to the side a little and the rear of the car will be brought back into the view of the mirror.

For a similar experience: I row, and occasionally use a cycling mirror to keep track of what's behind me -- that is, what's in front of the boat. When rowing (sculling) you face backwards. This occasionally presents navigational issues.

The mirror is almost completely useless. It's rather less useful on water than it would be on a bike.

The problem is orientation.

With a fixed mirror, you know where things are in relation to your vehicle (car, boat, bike, etc.).

With a head-mounted mirror, on a bike, you still have a generally fixed landmark to orient off of: the road, which over short distances exhibits to a good first-order approximation a high degree of linearity. That is: it goes before you, it goes behind you, you can draw a line between two points, and you have a pretty good idea of where things are relative to your position.

In a boat (dingy or rowing shell), there is no road (a straight marked rowing course might be different). Without a mirror, you generally navigate by occasionally checking over your shoulder to see if your dead reckoning is about right, then line up your stern (the back of the boat, but directly in front of your line of site). And you hope that nothing pops up suddenly in front of your bow (a buoy, a fixed marker, a swimmer, driftwood or debris, another boat, a breakwater, a dock, a pier, a sea lion, a cruise ship, a seagull, a submarine, a pelican, a whale .... and yes, I've encountered pretty much all of these at one point or another). You check over your shoulder at reasonably frequent intervals (every 10-20 strokes -- about 100-200 yards).

The mirror gives you some sense that there's something somewhere behind you.

It does a really crap job of telling you just where that something is.

If the field of view is large enough, you might be able to orient off the bow of the boat, but generally it's not and you're not fully stable enough to take advantage of that.

I generally row without the mirror.

And I keep my car's side-view mirrors adjusted so I can just see the sides of my car on the far inside of the mirror, to better judge positions of other vehicles relative to mine, blind spot (US flat mirror standards) be damned.

It also helps a little with reverse parking.

Yes, while driving, use cars in the lanes beside you. As a car passes you can adjust your side mirror so it catches the car just as it leaves your rear view mirror. With this setting, the passing car will enter your peripheral vision before it leaves your side mirror.

Once set you can find a head position that allows you see your own car, and you can use that position to reset your mirrors after the wife drives your car.

And that's why you should still do shoulder checks.

I use this mirror set up, but generally without over-the-shoulder checks. I can say that I've nearly hit people while changing lanes on multiple (but infrequent) occasions, when they were originally two lanes away from me, not visible in my mirrors, going slightly faster than me, and then changed lanes to the one beside me. They seem to sneak in during the second or two between when I first look in the mirror and begin the maneuver. It's a rare occurrence, but not rare enough for my comfort. I have my turn signal on the entire time (I start signalling when I intend to turn), and this still happens to me. I'm definitely going to start using over-the-shoulder checks again.

Pretty much all cars where I live (Finland) already have non-flat mirrors.

The most common type of mirror on the driver's side in new cars seems to be one which has a flat part (to show objects in their "true size") and a curved part at (to show a wider field of view).

No, because there is prior art. I'm not even sure it would stand in the US.

It's been a while since I have not driven a car that has a flat mirror near side (cost cuts and relative driver position makes far side less of a problem. Still, they're often not exactly flat either).

Cars here (Europe) generally have an improved version, as the linear radius was giving too much of a fisheye, which breaks depth linearity. The improved version starts with a quasi-flat mirror for the 2/3rd near the car, while the remainder has an increasing radius (Here's a shot [0]). The result is (when properly seated and mirrors set up) basically enough for the passing car to pop up at the edge of my field of vision when it disappears from the mirror.

[0] http://www.cleanmpg.com/photos/data/523/2007_Civic_iCDTi_Hat...

According to its PCT entry [1], it's still making its way through the Australian, Canadian, European, and Japanese patent offices.

1: http://patentscope.wipo.int/search/en/detail.jsf?docId=WO200...

US patents are only valid in the US. He has to obtain a patent in each country separately (other than in the European union where it is possible to get a "European Patent").

Curved mirrors are more efficient in one sense but tends to lead people to misjudge distances. Presumably the government think that showing a misleading picture may be worse than not showing one at all, which is not complete idiocy. This mirror apparently doesn't really suffer from that problem, but the legislation didn't anticipate it - which isn't really surprising, it's hardly realistic to expect them to foresee this development.

Reserve your judgment on government idiocy until we see how long it takes the government to update its legislation in response to this new invention.

It's not necessarily that a curved mirror is always worse, there's value in consistency so that drivers don't have to adjust to different cars. If there isn't a lot of value in picking the "right" answer, it may be beneficial for government to mandate one standard "good enough" solution.

I wouldn't be surprised if the original intention was to keep carmakers from putting shitty warped mirrors on the car.

Whatever, at least regulations have mellowed out a bit from the days when you were supposed to have someone walking a ways ahead of you to warn people a car was approaching.

It's not actually flat:

Hicks noted that, in reality, the mirror does not look like a disco ball up close. There are tens of thousands of such calculations to produce a mirror that has a smooth, nonuniform curve.

I imagine "tens of thousands of such calculations" is newspaper code for integration.

If someone assigned me the task of making such a mirror, I would start by treating the mirror like a monitor displaying a 3D scene, and calculate the vectors necessary to project a 45 degree view onto the surface of the mirror (i.e. start with a virtual camera behind the mirror such that it has a 45 degree field of view). I would then adjust those vectors so that the image will look like such a projection when seen by a viewer a few feet up and to the right, with the mirror occupying ~15 degrees of the viewer's FOV.

In other words, everything I need to know I learned from video games.

yeah. The ideal approach is a flat mirror with a curved one inset either above or below. That way you beg both.

However it seems to me that there is a bit of basic math at work here. The more you fit into a mirror of the same size, the smaller the objects will appear and the further away they will look. That strikes me as a very basic premise and I don't think you can get away from that so easily.

Passenger side mirrors are far and away the norm in the US.

I have never seen a car without a passenger side mirror.

> Even if it was, we don't typically give out patents for mere solutions to novel math problems.

Actually, we do give out patents for solutions to problems that involve math (the soluions involving math). Those patents do not cover using the same math for other problems.

Yeah, I was talking about idea patents.

Being able to patent an actual implementation is somewhat ok, but being able to patent ideas is plain stupid, and they are in the prrocess of importing it to the EU

You're right about that, but I don't usually find myself in such situations too often, and when I do I just move my head to get a better view straight back (either side/rear). I find blind spots much more dangerous :D.