Like many great Internet discoveries, Phillipe Cantin's journey started at Stack Exchange. He and his fellow LEGO fans have been working together in a beta forum about the toy brick, hoping the space would attract an audience and stick around. Last fall, Cantin posted that he wanted more insight from a Daily Finance article about why LEGO items were so expensive. The piece cited the use of acrylonitrile butadiene styrene (ABS) in LEGO bricks, a plastic that "requires petroleum as a raw material, [so] the cost of the plastic closely tracks the price of oil." LEGO, the article continued, feels compelled to use ABS due to its strength and resilience, two prominent factors in a brick's "clutch power"—the ability for pieces to snap together tightly while simultaneously being easy to separate.
That's where the question came in. "Since a LEGO brick is ABS plastic and will wear with use," Cantin wrote, "I wonder how many times you can assemble and dis-assemble two pieces before they no longer stay together." A few responses came in immediately, but they were mostly anecdotal. Cantin admitted that unless data was already available, he had a plan all along: build a robotic rig to determine the maximum "clutch power."
Cantin took to his workshop. He had to scrap his first design entirely when a metal tube attaching a servo-motor to a LEGO brick proved too flimsy (it died during initial testing). Ultimately, though, his final design only took two hours to conceptualize and build.
The entire setup is described in the video above (and shown in motion around the 2:45m mark). Cantin opted for an old CPU heat sink and, using an angle cutter, turned that into his new robotic arm. The rig uses the CPU heat sink arm to lift and drop the upper brick for the experiment. A servo-motor (running on a separate, nine volt power supply) holds and then releases the bottom brick to test "clutch power." There's an infrared sensor—which Cantin built himself—that determines whether the bottom brick does indeed lift and remain attached. Finally, an Arduino Uno controls the entire operation while logging each impression to a microSD card.
In his introduction video, Cantin said he was "not very happy with the engineering" for his ultimate rig but was "anxious to make something happen" with the testing. He felt the setup was too slow since one iteration took about 10 seconds (so ultimately, building it took way less time than using it). It was also potentially weak in his eyes. Cantin felt the hardware would fail well before the LEGO bricks did. Finally, the whole rig created a lot of commotion. "Don't try this at home," he warned on his blog. "It's long, noisy, and cruel to LEGO bricks." Cantin ran the whole thing anyway and, 10 days later, had his answer.
It took 37,112 impressions for the LEGO bricks—in this test alignment—to lose "clutch power."
“My daughter had a hard time sleeping because she could hear the robot even down in the basement," he told Ars. "I was happy when it finally stopped.”
Cantin's work ended in late February, but it caught the eyes of Arduino's official blog this week. With the added exposure, fellow LEGO fans had plenty of questions. Cantin released the video below to reveal further experiment details. He used a pair of 2x3 LEGO bricks because the piece is among the most basic and recognizable LEGO blocks (plus it has a better studs-to-weight ratio than say a 2x4 rectangle or the smaller squares). And this particular test used pieces from the 1970s, but that's because it was step one in Cantin's larger idea.
Cantin didn't think his initial experiment would take this long so he originally prepared to do nine tests in total: three using blocks from the 1970s, three with pieces from the 1990s, and three with current-day bricks. "I figured I'd do all my tests and plot that on a nice graph, because that would be something more meaningful statistically," he said. "But after the first test I realized I have to build a new rig all together. First of all, it's too long. And I didn't like the way the blocks were being assembled by the motion of the rotation; it's not a natural way to put LEGO bricks together. Pushing down in a straight matter makes more sense and maybe it's going to wear the bricks differently."
Cantin still has plans to move forward. "I like that number—37,000—but I want to see the answer." Regardless, he found value in the initial experiment. Friends estimated a LEGO brick would only last 400 to 500 impressions, Cantin himself thought the number would be more in the area of 5,000. "But there's no way a LEGO could be worn from this test. Cracking them or stepping on them is how they get broken."
Cantin is a Canadian senior software developer and this isn't his first foray into LEGO or Arduino. With LEGO, Cantin said he's been using them since he was young and never grew out of it. He worked with his young son to add working lights to the child's favorite LEGO robot and hopes to work on a tabletop robot game in the future. Previously, Cantin even used a similar Arduino-LEGO combination to responded to another LEGO Stack Exchange question. (Forum posters wanted to know if LEGO city-scale wheels could be motorized. Short answer: yes.) But if others were interested in recreating this particular experiment, he believes it's very accessible. All it takes are basic servo-controls, the SD logger available in various online libraries, and infrared detectors that can be purchased for the less hardcore.
If and when Cantin continues his research, you can be sure to find results popping up on the LEGO Stack Exchange and possibly the Arduino blog. But so far, no word from LEGO. Cantin does keep an eye on the Google Analytics for his site though, and unofficially he believes some traffic has come from the LEGO powers-that-be. "It would be nice if they contacted me, I would love that," he said. "To me, it'd be a childhood dream."
Listing image by flickr user: Molinary
reader comments
72