Original image

Zombies, Fire Drills, and Bad Decision-Making

Original image

Imagine that the dead have risen from their graves. They’ve gotten into a building you’re hiding out in. You slink down the hallway and enter what you think is a safe room.

It’s empty, and looks like a good place to hide. As you stand in the middle of the room, you look around. There’s only two doors: the one that you came through and one on the opposite side of the room. You should be able to barricade them both with furniture. But, oh no! The zombies have found the room, too. They’re shambling around both doors, with more crowding the doorway you just used, and now you have to get out. Which door do you exit through?

The less crowded one, I’m sure you’re saying. Of course, that makes the most sense. If both doors are the same distance from where you’re standing, why not use the one that’s got fewer obstacles?

Well, science has some bad news for you: You’ll probably wind up as a snack for the living dead, or at least stuck in a crowded doorway. Stress makes us do stupid things, like seek familiar routes even if they’re not the best ones. Over and over, eyewitness reports from real-life evacuations have suggested that, in emergencies, people tend to exit buildings from the main entrance that they used to enter the building, ignoring one or more emergency exits along the way. The crowding at these entryways slows evacuation times and sometimes results in injuries and even deaths.

Earlier this year, the Science Museum of London held a zombie-themed science festival called “ZombieLab.” Researchers Nikolai Bode and Edward Codling, from the University of Essex, took advantage of the event to look at the decisions people make in emergencies. They set up a computer simulation of a room evacuation similar to what I described above. One hundred and eighty-five museum guests took control of a computer person in a virtual environment filled with 80 virtual zombies.

At the start of the experiment, the participants just had to move their person from the hallway and into the central room. Next, they had to move back out again, through one of two doors, to where they started in the hall. During this second part, the researchers presented the visitors with a few different conditions. Some just had to simply exit the room. Others were encouraged to beat the fastest exit time. Others were presented with a crowd of zombies split unevenly between the two exits. A last group had to deal with the crowed exits while trying to beat the best time.

In the normal exit scenario and when they were trying to set the best time, the museum visitors split evenly between the two exit routes and showed no clear preference for one or the other. Faced with zombie-crowded exits, though, the visitors started to show some bias for the doorway they had come through, even if it was more crowded. Presented with the zombie obstacles alone, some of the visitors went for they door they came through, and then changed their mind when they realized how crowded the doorway was. With the added pressure of the time clock, fewer people changed their mind and stuck with trying to get out that exit, even though it was the slowest route.

Bode and Codling’s results fit with what other researchers have found in theoretical models and real-life evacuations. Under stress, people make irrational decisions. Here, the museum visitors under pressure to exit quickly were more likely to stick to the route they knew even if it wound up taking them longer to get out, and were less likely to change their mind and adapt to the situation.

In a real-world situation, the researchers say, their results suggest some strategies for minimizing risks during stressful evacuations. One idea they offer is having people in large, crowded buildings enter from several different locations. If they have to get out quickly, and their preference for the way they came in holds up, they’ll spread out to different routes and avoid overcrowding any one exit.

It’s worth mentioning that the idea that the other virtual characters in the room were zombies was just meant to fit the experiment in with the theme of the festival and keep participants blind to the purpose of the experiment. In the simulation, the zombies didn’t attack participants or pose any danger, but simply blocked the doorways. The study participants didn’t have to treat them as a threat, so they focused on choosing one door or the other without worrying about getting their brains eaten. I wonder if, or how, the results would differ if the “zombies” acted more like zombies, and how decision-making in an evacuation is affected if there are obstacles at exits that pose active threats.

Original image
iStock // Ekaterina Minaeva
Man Buys Two Metric Tons of LEGO Bricks; Sorts Them Via Machine Learning
May 21, 2017
Original image
iStock // Ekaterina Minaeva

Jacques Mattheij made a small, but awesome, mistake. He went on eBay one evening and bid on a bunch of bulk LEGO brick auctions, then went to sleep. Upon waking, he discovered that he was the high bidder on many, and was now the proud owner of two tons of LEGO bricks. (This is about 4400 pounds.) He wrote, "[L]esson 1: if you win almost all bids you are bidding too high."

Mattheij had noticed that bulk, unsorted bricks sell for something like €10/kilogram, whereas sets are roughly €40/kg and rare parts go for up to €100/kg. Much of the value of the bricks is in their sorting. If he could reduce the entropy of these bins of unsorted bricks, he could make a tidy profit. While many people do this work by hand, the problem is enormous—just the kind of challenge for a computer. Mattheij writes:

There are 38000+ shapes and there are 100+ possible shades of color (you can roughly tell how old someone is by asking them what lego colors they remember from their youth).

In the following months, Mattheij built a proof-of-concept sorting system using, of course, LEGO. He broke the problem down into a series of sub-problems (including "feeding LEGO reliably from a hopper is surprisingly hard," one of those facts of nature that will stymie even the best system design). After tinkering with the prototype at length, he expanded the system to a surprisingly complex system of conveyer belts (powered by a home treadmill), various pieces of cabinetry, and "copious quantities of crazy glue."

Here's a video showing the current system running at low speed:

The key part of the system was running the bricks past a camera paired with a computer running a neural net-based image classifier. That allows the computer (when sufficiently trained on brick images) to recognize bricks and thus categorize them by color, shape, or other parameters. Remember that as bricks pass by, they can be in any orientation, can be dirty, can even be stuck to other pieces. So having a flexible software system is key to recognizing—in a fraction of a second—what a given brick is, in order to sort it out. When a match is found, a jet of compressed air pops the piece off the conveyer belt and into a waiting bin.

After much experimentation, Mattheij rewrote the software (several times in fact) to accomplish a variety of basic tasks. At its core, the system takes images from a webcam and feeds them to a neural network to do the classification. Of course, the neural net needs to be "trained" by showing it lots of images, and telling it what those images represent. Mattheij's breakthrough was allowing the machine to effectively train itself, with guidance: Running pieces through allows the system to take its own photos, make a guess, and build on that guess. As long as Mattheij corrects the incorrect guesses, he ends up with a decent (and self-reinforcing) corpus of training data. As the machine continues running, it can rack up more training, allowing it to recognize a broad variety of pieces on the fly.

Here's another video, focusing on how the pieces move on conveyer belts (running at slow speed so puny humans can follow). You can also see the air jets in action:

In an email interview, Mattheij told Mental Floss that the system currently sorts LEGO bricks into more than 50 categories. It can also be run in a color-sorting mode to bin the parts across 12 color groups. (Thus at present you'd likely do a two-pass sort on the bricks: once for shape, then a separate pass for color.) He continues to refine the system, with a focus on making its recognition abilities faster. At some point down the line, he plans to make the software portion open source. You're on your own as far as building conveyer belts, bins, and so forth.

Check out Mattheij's writeup in two parts for more information. It starts with an overview of the story, followed up with a deep dive on the software. He's also tweeting about the project (among other things). And if you look around a bit, you'll find bulk LEGO brick auctions online—it's definitely a thing!

Original image
Opening Ceremony
These $425 Jeans Can Turn Into Jorts
May 19, 2017
Original image
Opening Ceremony

Modular clothing used to consist of something simple, like a reversible jacket. Today, it’s a $425 pair of detachable jeans.

Apparel retailer Opening Ceremony recently debuted a pair of “2 in 1 Y/Project” trousers that look fairly peculiar. The legs are held to the crotch by a pair of loops, creating a disjointed C-3PO effect. Undo the loops and you can now remove the legs entirely, leaving a pair of jean shorts in their wake. The result goes from this:


Opening Ceremony

To this:


Opening Ceremony

The company also offers a slightly different cut with button tabs in black for $460. If these aren’t audacious enough for you, the Y/Project line includes jumpsuits with removable legs and garter-equipped jeans.

[h/t Mashable]