CLOSE
Original image
YouTube / ONE

How Missed Calls Amplify Farmers' Voices

Original image
YouTube / ONE

This week, Farm Radio International (FRI) announced the results of an innovative poll covering thousands of farmers. The biggest surprise was the way farmers voted: by calling a phone number and hanging up.

The survey was conducted in Tanzania, where smallholder farms (small family farms) make up around 75% of all farm production. FRI, an international radio service that partners with local stations, wanted to poll those farmers in order to help make their voices heard by the Tanzanian government. But how do you reach thousands of tiny farms spanning a whole country? In the case of Tanzania, the answer was radio talk shows and basic cell phones.

Photo courtesy of ONE / Do Agric

The Power of Radio Talk Shows and Cell Phones

Across Tanzania, there are radio stations broadcasting talk shows aimed at farmers. Those programs are already popular for the people the survey aimed to reach, so FRI partnered with five radio stations in different regions across the country. The local presenters added discussion segments to their programs dealing with the poll issues.

Radio broadcasters concluded the poll segments by asking yes/no questions, then giving out phone numbers that voters could dial into. But people generally don't want to waste their cell phone minutes on a poll, so a clever solution came into play: just call the number, then hang up. The missed call is logged, and that log constitutes a vote. This system is called "Beep to Vote," and it's free for voters because the missed call doesn't incur charges for using cell phone minutes. For yes/no questions, there was one phone number for "yes" and another for "no." A total of 8,891 smallholder farmers participated.

In addition to the "Beep to Vote" yes/no questions, the poll included a multiple-choice question that most voters responded to using SMS. Voters texted a single character ("1" for the first option, "2" for the second, and so on) to a specified phone number, and those results were tallied by computer. In addition to the SMS voting method, farmers could opt to make a voice call to an automated system, listen to the five options, and press a number to indicate their choice. 4,372 people responded to the multiple-choice question. The system was also able to send SMS reminders to voters in case they voted for one of the poll questions, but not the others.

The data was crunched in realtime using a system made by Telerivet, so poll workers could watch as votes came in. The system also checked incoming phone numbers so each phone (which roughly equates to each voter, or household) could only vote once per question.

Photo courtesy of ONE / Do Agric

Why This Matters

From a technological perspective, this poll is a brilliant example of choosing the right technology for the job. If a similar poll were conducted targeting middle-schoolers in the United States, it's likely that technologies like YouTube videos and click-to-vote within the video would be used. But for these Tanzanian farmers, the prevalent technologies are radio and cell phones. By putting them together, in a near zero-cost way, FRI was able to collect data that could influence government policies, which in turn could change livesusing just cellphones and radio.

This poll was part of a campaign called Do Agric, focused on encouraging African leaders to invest more in agriculture, in order to improve farming (and in turn, daily life) in Africa. Here's a video about the program:

When the results were announced earlier this week, Tanzania's President Kikwete said, "Action on agriculture has to be today, not tomorrow!" The voices of 8,891 farmers reached the president's ears.

For more on the survey, check out FRI's page on methodology and results.

Original image
iStock // Ekaterina Minaeva
technology
arrow
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
Stephen Missal
crime
arrow
New Evidence Emerges in Norway’s Most Famous Unsolved Murder Case
May 22, 2017
Original image
A 2016 sketch by a forensic artist of the Isdal Woman
Stephen Missal

For almost 50 years, Norwegian investigators have been baffled by the case of the “Isdal Woman,” whose burned corpse was found in a valley outside the city of Bergen in 1970. Most of her face and hair had been burned off and the labels in her clothes had been removed. The police investigation eventually led to a pair of suitcases stuffed with wigs and the discovery that the woman had stayed at numerous hotels around Norway under different aliases. Still, the police eventually ruled it a suicide.

Almost five decades later, the Norwegian public broadcaster NRK has launched a new investigation into the case, working with police to help track down her identity. And it is already yielding results. The BBC reports that forensic analysis of the woman’s teeth show that she was from a region along the French-German border.

In 1970, hikers discovered the Isdal Woman’s body, burned and lying on a remote slope surrounded by an umbrella, melted plastic bottles, what may have been a passport cover, and more. Her clothes and possessions were scraped clean of any kind of identifying marks or labels. Later, the police found that she left two suitcases at the Bergen train station, containing sunglasses with her fingerprints on the lenses, a hairbrush, a prescription bottle of eczema cream, several wigs, and glasses with clear lenses. Again, all labels and other identifying marks had been removed, even from the prescription cream. A notepad found inside was filled with handwritten letters that looked like a code. A shopping bag led police to a shoe store, where, finally, an employee remembered selling rubber boots just like the ones found on the woman’s body.

Eventually, the police discovered that she had stayed in different hotels all over the country under different names, which would have required passports under several different aliases. This strongly suggests that she was a spy. Though she was both burned alive and had a stomach full of undigested sleeping pills, the police eventually ruled the death a suicide, unable to track down any evidence that they could tie to her murder.

But some of the forensic data that can help solve her case still exists. The Isdal Woman’s jaw was preserved in a forensic archive, allowing researchers from the University of Canberra in Australia to use isotopic analysis to figure out where she came from, based on the chemical traces left on her teeth while she was growing up. It’s the first time this technique has been used in a Norwegian criminal investigation.

The isotopic analysis was so effective that the researchers can tell that she probably grew up in eastern or central Europe, then moved west toward France during her adolescence, possibly just before or during World War II. Previous studies of her handwriting have indicated that she learned to write in France or in another French-speaking country.

Narrowing down the woman’s origins to such a specific region could help find someone who knew her, or reports of missing women who matched her description. The case is still a long way from solved, but the search is now much narrower than it had been in the mystery's long history.

[h/t BBC]

SECTIONS
BIG QUESTIONS
BIG QUESTIONS
JOB SECRETS
QUIZZES
WORLD WAR 1
SMART SHOPPING
STONES, BONES, & WRECKS
#TBT
THE PRESIDENTS
WORDS
RETROBITUARIES