Three Months, $6,500, and Billions of Hours of Fun
In 2.009, or Product Engineering Processes, students generate thousands of new product ideas and turn a few really special ones into full-fledged prototypes. Technology Review follows one student team—chosen at random—on its 2.009 journey.
September 7, 2011. It’s a few minutes past 1:00 p.m. as David Wallace, SM ’91, PhD ’95, paces before the wall of blackboards in 10-250, surveying the crop of 101 students who have enrolled in 2.009, Product Engineering Processes.
The mechanical-engineering professor wears his usual uniform: cargo shorts, steel-toed work shoes with dark socks, and a T-shirt. His pacing is the only sign of the nervous energy that lurks beneath his low-key demeanor. At 1:08, he begins by demonstrating prototype products from past years. Then he springs the semester’s first mini-quiz: which of several shapes is different from the others? As it turns out, there’s more than one answer.
In product engineering, Wallace tells the students, there can be a lot of right answers, like a bike designed for a girl with cerebral palsy that got her riding in just a few minutes. But there are even more wrong answers, because so many little details—like choosing unsuitable O-rings for the space shuttle—can spell disaster. “It’s empowering, building the environment we live in,” he says. “But it’s a big responsibility, because there are so many ways things can go wrong.”
The students soak it all in. They know they’re in for a semester that will be grueling and, for many, incredibly rewarding. Although it’s officially a 12-hour class (three hours in class, three in the lab, and six hours of homework each week), students routinely spend 20 hours a week on 2.009—and decidedly more during crunch times. (The TAs joke that it adds up to “100 billion person-hours of fun.”) The course, originally known as 2.73, dates back at least to the 1950s. For decades, students designed on paper only; by 1995, when Wallace took over the class, they were building their projects. Under Wallace, it has evolved to give mechanical-engineering students (who are required to take it) and a few curious souls from other courses a realistic taste of life as a product design engineer.
2.009 course website
2.009's 2011 photo gallery
2.009 class videos
They work in big teams, keep detailed design notebooks, offer feedback to their peers, log their hours on time sheets, and create a fully functioning product prototype by semester’s end. The teams work through six critical milestone reviews along the way, culminating in an Apple-style product launch in Kresge Auditorium.
Wallace, the recipient of several MIT teaching awards, puts in at least as much time as the students to make the class what is, for many, a defining experience of their MIT career. When he surveyed students in 2004 about their career plans before and after they took 2.009, the appeal of management consulting and medicine declined dramatically, while interest in product development and grad school shot up. “It’s the class you look forward to when you sign up for Course II,” says Stephen Hendel ’12.
September 9. The buzz in the hallway in front of 10-250 escalates as Wallace and TAs Melody Kuna ’11, Ilan Moyer ’08, and Josh Ramos ’11 ready the classroom for what he calls the “Big Reveal.” Today, he’ll tell students the theme around which they’ll design their products—a theme drawn from a short list after a July brainstorming meeting of around 20 of the instructors, mentors, and TAs who help run 2.009. As a former ice dancer who competed at the national level in Canada, Wallace understands the value of showmanship, and he has cooked up a memorable way to announce the theme.
He begins by talking about the shape quiz from the first session. Most students had assumed there was one right answer. “Nothing is more dangerous than an idea when it’s the only one we have,” he says, quoting the philosopher Emile Chartier. He adds, “If you’re working on an obvious solution, probably your competitors will be working on it, too.”
In an idea generation exercise he threw at the students in the first class, they averaged 1.2 ideas per minute, which he says “borders on mental constipation.” He tells them to shoot for four ideas per minute by looking for the next right answer, looking at things from multiple viewpoints, and deferring all judgments.
Maria Yang ’91, an assistant professor of mechanical engineering and engineering systems, suddenly rolls in, riding a toy pony on casters. Wallace asks whether her pony bites, and she says no. He reaches out to pat the pony, then draws back his hand with a yelp of pain. “I thought you said your pony didn’t bite!” he accuses, rubbing his hand. “It’s not my pony,” Yang replies. Wallace uses this classic gag to relay another creativity strategy: challenge assumptions.
Working up to the introduction of the theme, Wallace tells the students they are part of a product development firm with a long history of innovation. He hits a remote to start a musical fanfare. But instead, the name of the theme—”On the Go”—pops up prematurely on the classroom screen. He acts as if nothing’s wrong, and in a split second he decides to ditch the fanfare. He exits the classroom and bounds back moments later as the embodiment of exercise guru Richard Simmons—corkscrew wig, sparkly tank top, silky shorts, and all.
Wallace, who has even shaved his trademark beard to play this role, milks it for all it’s worth as he and a tuxedo-wearing Moyer pull back a red cloth to reveal products that seem to shimmer with promise through a dry-ice fog. The theme is a starting point: ”Simmons” says the students can design products relating to such fields as transportation, food, medicine, mobility and assistive technology, travel, and recreation. Each of the eight teams of 11 to 13 students will have $6,500 and three months to zero in on an idea and develop a fully working prototype. The class is officially under way.
September 12. Wallace lopes gracefully across Killian Court, megaphone in hand. “Five minutes up, five minutes to go!” he shouts, as the newly formed 2.009 teams race to build the tallest free-standing balloon structure they can using 144 balloons in their team’s color and three rolls of packing tape.
The teams are assembling for the first time, but Wallace has designed the exercise to be more than a mere ice-breaker. Before the contest started, each student drew a slip of paper assigning a role to play: “You are very concerned about every detail … You work slowly but what you do is done well.” “You have little patience for others.” “You don’t really care about the exercise; you’re just here to have as much fun as possible.” The Red Team’s structure towers several feet above the tallest student until the group’s “have fun” student sneaks up and pops it with a pen. When time’s up, Wallace declares the Orange Team the winner of a Toscanini’s gift card.
As the teams stand around and discuss their roles, Wallace confesses to being perpetually amazed that they do stand around and talk about the balloon challenge. He thinks it’s better than meeting each other for the first time around a table in the lab. With the sun shining on a leafy Killian Court filled with balloons and chatting students, no one seems to disagree.
September 15. At 9:00 a.m. the Orange Team assembles around the Pappalardo Lab’s conference table in the basement of Building 3 for its first official lab meeting. Lab instructors Maria Yang (who does not, in fact, own a pony) and Juhan Sonin (creative director for a product design firm) introduce themselves. Orange Team mentors include three grad students, two Lincoln Laboratory engineers, communications instructor Atissa Banuazizi, senior lecturer Dan Braunstein SM ’94, PhD ’98, and industry mentor Robin Miller, a product technical leader at Pratt & Whitney.
But as Wallace made clear in the weekly 2.009 staff meeting, instructors and mentors are not running the show; they instead serve as questioners, guides, role models. (“At first, you’re like Obi Wan, asking teasing questions,” he told them. “Later, you’re a helper, holding something while they work on it.”) The students may spend 25 percent of their professional lives in meetings; Wallace wants them to learn how to run effective ones. (Wallace himself hasn’t had to endure many corporate meetings; he had a four-month internship at Philips and one other corporate job before quickly heading to academia, where he says he can do cutting-edge things and take more risks.)
After introductions, the team splits into two sections. Orange A and Orange B then set about choosing a systems integrator from their own ranks—someone who will run the meetings and organize the work to be done. Orange B resorts to rock-paper-scissors to determine who will have the first try at systems integrator; Nydia Ruleman ’12 wins. After divvying up the remaining team roles (financial officer, wiki master, information officer, tool officer, safety officer), they discuss the results of their brainstorming assignment. Each student had to come up with at least 20 ideas in a design notebook, sketching out five in detail. They take turns presenting their contenders, pinning them up on the movable wall that now divides the conference room.
Senior Jessica Iacobucci’s ideas include a battery-operated hair straightener and portable furniture that wraps around a tree. Charlie Klene ’12 has dreamed up boots with retractable spikes and a pocket defibrillator. David Parell ’12 came up with a tote that charges a battery as you swing it and what he calls the Omnibrush, a mouthguard-like device with bristles that lets you brush all your teeth at once. Bennett Wilson ’12 has sketched a shopping cart optimized for liquor stores and a Swiss Army knife–style travel device that holds a toothbrush, toothpaste, a comb, and a clock.
Ruleman’s ideas include adjustable ice wraps for athletes and day-to-night shoes with folding, retractable, or snap-on heels. Nick Dou ’12 presents an antitheft wallet and a bike lock that beeps if you try to cut it. They begin sorting through their 30 sketches, looking for three worth presenting to the whole class. As the three-hour lab wraps up, they decide to spend more time generating ideas and meet again Sunday afternoon.
September 19. In Monday’s class, Wallace preps the students for their first major milestone, the three-ideas presentation on Friday. The exercise, like all 2.009 milestones, will let instructors offer critical feedback to help the students make better decisions as they home in on a product and then refine it. Each lab section will give three elevator pitches in two minutes; when the timer goes off, the figurative elevator door will shut.
After the advice comes an in-class design challenge. In past years Wallace, who for a time helped design the engineering challenges for the PBS kids’ show Design Squad, has asked 2.009 students to devise ways to walk on water and to set a ship on fire using mirrors, emulating Archimedes’ death ray.
This year, he taps into the popularity of the game Angry Birds. He gives the students 10 minutes to design a better launching device to help three-dimensional representations of the on-screen birds catapult themselves onto evil pigs who have stolen their eggs. Pencils begin scratching—or remain aloft as their owners spend precious minutes pondering. Although they don’t know it yet, the students’ back-of-the-envelope ideas will soon take on a life of their own.
MILESTONE 1: THREE-IDEAS PRESENTATION
September 23. A height-adjustable scooter. Origami-inspired shelters for the homeless or displaced. An electromagnetically secured fire-truck hose. A collapsible walker. Headphones with gesture controls. A vending machine that dispenses bike helmets. Chair Force One, a wheelchair that electrically boosts the user’s manual force. The Forget-Me-Not, a mat on which to place easily forgotten items; it beeps if the user walks out the door without them. In 32 minutes, the 16 lab sections present 48 ideas.
In a few presentations, the product itself or the technology behind it is not well explained, and some ideas leave the instructors shaking their heads. Crutches with wheels? An alarm clock that brews coffee? And perhaps the need for an automatic peanut-butter-and-jelly-sandwich maker isn’t quite as compelling as its extremely enthusiastic presenter suggests (“It’s peanut butter and jelly time!”).
But only a few sections get cut off by the two-minute buzzer, and several polished, lucid ideas stand out. Yellow A’s Marcel Sanchez ’12 asks for a show of hands by those who floss daily, then reports that non-flossers are at higher risk for heart disease. He’s got everyone’s attention as he says that a device he calls Brush ‘N’ Floss would let users floss their teeth with a jet of water as they brush. His teammate Sam Powers ’12 pitches Run-Run Revolution, a game involving pressure sensors and accelerometers in the player’s shoes.
The Orange Team’s presentations go well. Orange A’s Jared Darby ’12 proposes portable tables and stools for the Clover food truck. Kami Klauber ’12 pitches Quick Cup, which would let users control the temperature of a hot drink. Richard Dahan ’12 describes Mobile Meds, a drug-sorting device for patients.
Orange B’s pitches by Klene, Parell, and Wilson include StepCase, wheeled luggage designed to go up stairs, and Wilson’s Travel MultiTool. Their presentations are decidedly smoother than the practice run-throughs in their Thursday lab meeting. Wilson’s “um”s have disappeared, and he gets laughs by drop-kicking his toilet kit across the stage during his MultiTool pitch.
September 26. Product design classes at other universities typically focus on developing a predefined product concept, but the mantra in 2.009 is “Ideate. Model. Test!” In fact, a surprisingly large chunk of the semester—roughly half—is devoted to idea generation. Wallace likens 2.009 to a reality TV show in which ideas are the contestants that get voted off the island each week.
For the next big milestone, the sketch model review, the teams will develop a portfolio of ideas related to a subtheme chosen by the instructors. They will then build at least four sketch models—simple physical models made of inexpensive materials—so they can test and analyze several ideas before choosing the one they’ll ultimately build. Making sure you’re working on a worthy problem, says Wallace, is “like showing up at the right hockey rink for your game.”
October 3. Wallace has the students recite two pledges: “I will work safely in the lab” and “I will not flip birds in the lab.” Then the class heads down to Pappalardo, where each team will use kits assembled by Wallace and the TAs to build one of the catapults that students sketched during the Angry Birds design challenge. “No flipping birds” means no testing the devices in the lab.
By the end of class, eight variations on the catapult, ranging from the modern to the medieval, line the length of Pappalardo. Orange has built an air cannon. Ancestors of Yellow’s bow-powered ballista, Red’s trebuchet, and Purple’s twisted-rope onager might have launched flaming torches, hot oil, or dead bodies bearing contagious diseases over castle walls in the Middle Ages. The students are itching to try out their handiwork, but they keep their pledge.
October 5. At 5:30 a.m., it’s still dark when Wallace and his three TAs assemble in Pappalardo to begin testing the catapults. They push Silver’s guided slingshot into the parking lot, Wallace paces off 20 meters, and using a stuffed ball as a stunt double for an Angry Bird, they take a shot. The ball sails a disappointing six meters before landing. They adjust the angle of the slingshot and send the ball a good 30 meters. The onager needs a stiffer rope and more twisting. The trebuchet flings the ball high in the air but not very far forward; Wallace prescribes more elastic ties.
The air cannon won’t hold pressure from a bicycle pump, the air compressor in the trunk of Wallace’s car, or even the lab’s heavy-duty air pump. The TAs coat it with soapy water to test for leaks in the glue joints. “It’s a little soap-bubble machine at the moment,” Wallace observes. The joints will need to be resealed, if there’s time. There seems to be little danger that any birds will be flung into Memorial Drive traffic when the class tries out the catapults in Killian Court in two days.
October 6. When the 13 Orange Team students gather for their Thursday lab, the progress they’ve made since the previous week is remarkable. Orange B met on yet another Sunday and decided to ditch all three ideas it had pitched to the class.
Zeroing in on the team’s assigned food subtheme, Iacobucci mentioned that she has an uncle who is a professional chef, which led her group to two ideas: Ultra Clean, a device that creates sonic waves in a sink to clean pots, and a sonar-based sensor that attaches to a faucet and lets users set the level to which a pot should be filled by placing a hand at that level on its outside. These fit the “On the Go” theme because they would free up kitchen staff to move around and attend to other tasks, like making food.
The team ordered parts on Monday and built the two devices on Tuesday. By Thursday’s lab, they have two rough prototypes. The pot filler is “really slow,” says Sonin, “but they’re proving the concept and can easily increase the flow rate later.” They devote most of the lab to preparing their presentation for the sketch model review, now just hours away.
On the other side of the conference room, Yang is telling Orange A’s presenters it’s important to show that they talked to potential Mobile Meds users. Bryan Macomber ’12 is putting the finishing touches on slides he’s created about Four Leaf, the portable table designed for the Clover food trucks; it’s made of four pieces that clamp around trees, telephone poles, or maybe even street signs.
MILESTONE 2: SKETCH MODEL REVIEW
At 7:00 p.m., the sketch model review begins. There’s a mix of old and new ideas, including a folding bike; Homey-gami, the portable folding shelter; a device (inspired by Mrs. Weasley’s clock in the Harry Potter books) that uses RFID to show the location of items in a house; TranSit, a seat that folds down from the side of a bus or subway car; and the bike helmet dispenser, now known as Helmet Hub.
Orange A’s Four Leaf presentation goes smoothly, but when Brent Boswell ’12 takes over the remote to introduce Mobile Meds, he can’t pull up the slides and is forced to wing the presentation. An Advil gets stuck in the plexiglass sketch model, and the six-minute buzzer sounds before Boswell finishes. Orange B fares better as Ruleman presents Ultra Clean, and Parell and Wilson talk through the sketch model of their automatic pot filler. The flow rate is so slow that Iacobucci speeds the demo by adding water to the pot manually. But the water flowing from the faucet stops at the right level, and Parell draws a laugh when he tells the audience the name of their device: Phil.
October 7. Wallace gives a detailed post-mortem of the sketch model review and shares the instructors’ and mentors’ project rankings. The Orange Team has good reason to be happy. Although Mobile Meds comes in near the bottom and Ultra Clean is in the middle of the pack, Four Leaf ranks fourth. And Phil tops the list.
Now Wallace shifts gears, asking the students to recite two more oaths: “I will be careful” and “I will flip birds in Killian Court.”
The class streams into the Friday-afternoon sunshine to find MIT’s famous courtyard transformed into a giant testing ground. Each team’s launcher awaits, ready to fling a nestful of cork-and-oatmeal-stuffed birds rendered in its team color against a target of cardboard structures topped with water-filled green balloons as stand-ins for the evil pigs. (Wallace designed the birds himself, buying an embroidery machine to fashion their signature scowls.)
The professor sounds the starting gun, and birds begin sailing across Killian Court. Orange’s air cannon could not be fixed in time, so the students rely on their own force to pull a big slingshot with two lengths of surgical tubing. Amid cheers and groans, Green finishes first and wins a giant Stanley Cup–like trophy made of blue foam.
Wallace, dressed as the chief green pig from the Angry Birds game, is hoarse after a week of getting little more than an hour of sleep a night as he and the TAs readied the catapults. “If people remember and get excited about doing this type of stuff,” he says, “it’s worth it.”
October 13. Orange assembles for lab on Thursday morning, having met the night before to talk through the options. The team has decided to focus on refining Phil and Four Leaf. They rejigger the lab sections that will develop each idea and the newly reconstituted Orange B gets to work on Four Leaf. What sort of clamp should support the table? Does it have to be a clamp? Using a blood-pressure cuff proved too slippery on a PVC pipe; perhaps a sticky material like rubber could help? But what if it’s punctured?
Sonin recommends getting details on the dimensions and materials of any kind of pole the table might attach to—telephone, lamp, signpost, tree trunk. “And have you looked at how people act on a street around a food truck?” he says. “Spend an hour at lunchtime and videotape it.” They draw up a list of constraints: one person should be able to set up the table in 30 seconds, using no tools. It should hold 100 pounds. It must weigh no more than five pounds.
Meanwhile, Orange A covers the conference room’s blackboard with a to-do list for Phil, organized in overlapping phases. The students must talk to potential users in kitchens, develop a CAD model, spec and order a valve and sensor, code the features, and more. Everything must be tackled in the eight days before the mockup review.
October 20. In Orange A’s morning lab meeting, it turns out that conversations with potential customers have revealed a surprising fact: the Italian restaurants the students had assumed were constantly filling pots to cook pasta actually reuse the water. Most professional chefs aren’t, in fact, hankering for an automatic pot filler. But the three-step process of rinsing, washing, and sanitizing dishes in a three-tub sink would be easier with an automatic sink filler.
On Orange B, Klauber and Darby head out to photograph Four Leaf on a pole near the student center. After watching them from afar, some MIT facilities workers on a coffee break amble over, eager to try the table. They’re enthusiastic about the idea because the steps of the student center, where they usually sit, can be wet with rain.
MILESTONE 3: MOCKUP REVIEW
October 20. Tonight each lab section gets three minutes to present its idea; then everyone moves to the lab for demos and Q&A sessions with rotating teams of instructors and mentors. Both of Orange’s sections end up high in the instructors’ rankings when the night draws to a close. Having two strong ideas is great, but it will make it painful to decide which one to leave behind.
MILESTONE 4: THE DECISION
October 27. Now Orange has to make that choice: should the team jettison Four Leaf or Phil? Before the debate begins, four students are in the Phil camp, five want to pursue Four Leaf, and four are neutral.
With Four Leaf, figuring out the clamping mechanism would be a good mechanical-engineering challenge, and the product could improve the experience of eating at a food truck. Having an enthusiastic potential customer in Clover is a big plus. The team could develop a prototype and see it put to use quickly. “It would be cool to walk around campus and say, ‘I made that!’” Iacobucci says. Cons: the clamping mechanism still needs to be designed, so there’s not a clear path to producing the table. Will a permit be required to use it? Will they be able to design it for one-person assembly? If someone sits on a table and it collapses, will the Orange Team be liable?
Phil has many pros: the next steps are clear, extensive user research has already been done, and it has a definite “wow” factor. The major con is that it will require a lot of electrical-engineering expertise, limiting the ability of many people on the team to contribute. But Cat Thu Nguyen Huu ’12 points out that Four Leaf would require work to be done in sequence; the table couldn’t be designed until the clamps were settled. With Phil, people could work in parallel. Eventually the tide turns for Phil.
When they reach a consensus, the students follow Wallace’s instructions to celebrate by performing a Japanese clapping ritual known as Tejime. With some giggles, the decision is sealed.
MILESTONE 5: ASSEMBLY REVIEW
November 7. In class, Wallace recaps the previous week’s assembly review, for which teams had to produce an assembly model, creating 3-D digital models (or solid models) detailing each part in their design. Teams also had to create a product contract spelling out their design specifications and criteria for measuring the success of the design. Details matter when you’re assembling a product, so parts, materials, and form all came under scrutiny—as well as ease of assembly, manufacturing, and use. Today, Wallace commends Orange for presenting not only complete fabrication details but also a well-targeted need, a specific value proposition, and a clear product vision; Thursday, they’ll be served breakfast at their lab.
Wallace then shows the class an alpha prototype of a new type of cleaner he says he has been working on. When he plugs it in, sparks fly, the device begins to smoke, and the professor looks embarrassed. He moves on, offering advice for the upcoming technical review: work in parallel, keep meetings short, communicate who’s doing what, design (don’t hack), and above all, test. “If it’s not tested, it does not work,” he says. During the mini-quiz that follows, two guys dressed as a gorilla and a huge banana wander into the classroom and deposit a toy monkey on the front table.
There is (as always) a method to Wallace’s apparent madness. Only a few students notice the gorilla and the banana; most are so focused on the quiz that they miss the scene. The lesson: be alert for unexpected things. Likewise, the exploding prototype also turns out to be a calculated lesson: Wallace roughed up an air purifier to make it look like a prototype and wired it to spark and emit smoke when he plugged it in. His point was that carefully observing and recording a problem and its symptoms make up the first critical stage of debugging. “Wishing will not make a fault go away,” he says. “Symptoms that mysteriously disappear are not problems solved.” There are no shortcuts to debugging—or to any aspect of designing a great product.
November 17. In lab, the Orange Team reviews plans for Phillionaire, as the students are calling the latest version of their prototype, Phil 6.0. (The five previous versions were called Bucket Phil/Phil Billie, Mockup Phil, Philippe, Phyllis, and Chill Phil.) Tasks on today’s agenda include placing the buttons, installing electronic components and progress-bar LEDs, and practicing drilling holes for the buttons and the battery. Macomber says they have to work to get the prototype looking less like a toilet. “Don’t worry about that now,” says Sonin. “Get it working. What rocks is that you’re at 6.0. Most groups only get one version.”
MILESTONE 6: TECHNICAL REVIEW
November 21. The technical review is all show and no tell. The instructors rotate through the lab, trying out the prototypes and firing questions at the teams to make sure they’ve thought through every detail. Orange had added a nozzle to Phil to improve its water flow that afternoon, but it interferes with the sonar, so the first demo doesn’t work. “We learned a valuable lesson,” Dahan will say later. “Don’t change something two hours before a demo. Our 15 hours of testing went down the tubes.” To make matters worse, a poker game at Orange’s lab table—stress relief after several all-nighters—raises classmates’ and instructors’ eyebrows.
November 30. After class, Wallace admits that he’s nervous for some of the teams. “This is show time. You can’t fake it in the end,” he says. “You can’t work in pretend mode and then expect to be good when it counts.”
As the teams’ work ramps up, Wallace himself is seeing how little sleep he can live on for three weeks as he undertakes the huge number of preparations for the final presentations: designing and building professional sets for each team, overseeing elaborate AV setups, scripting music and videos that will punctuate the presentations to make sure the evening doesn’t drag. The three TAs are also putting in crazy hours to produce videos from the roughly 30,000 photos they’ve taken throughout the term; the technical staff at the Pappalardo Lab will individually log between 85 and 103 hours in the final week alone.
“You can’t ask people to work harder or put in more than you’re willing to do yourself,” says Wallace, whose commitment to continuous improvement leads him each year to choose the one or two least effective 2.009 lectures and completely revamp them. “Students get that and respond. Every single thing you do is an example.” Wallace may be setting an extreme example, but given the students he’s teaching, he thinks that’s appropriate. “MIT’s mission is to send people off to do the special things, the things that will change the world,” he says. “This class is about trying to light a fire and helping people appreciate the excitement of pushing, of doing something as well as you can.”
December 5. With the final presentation a week away, today’s class meets in the lab. Orange will produce the parts for Phil 7.0 on a 3-D printer tonight. Boswell reports that the faucet attachment is sticking. Darby investigates ways to get Phil to look like stainless steel instead of plastic. Iacobucci works on brochures for the final presentation. Dou and Ruleman shoot video of the Sloan Café kitchen for the presentation. Macomber designs the slide show. Klauber does patent research. “We’ve reached the level where we’re now already at good enough,” says Dahan. “The only question is, can we go from good enough to fantastic?” The buttons, for example, are not waterproof, and the team is working on what he calls “a more legit progress bar.”
There’s no time for much of anything else in the teams’ lives, even with finals looming. As one student (who wants to remain nameless to avoid offending other professors) puts it, “I only have one class, and it’s called 2.009.”
December 8. As Orange hammers away on its presentation and marketing plan, Wallace (wearing a T-shirt that says “Make Things”) and the shop staff build sets. One shop technician is cutting window holes out of the front of Orange’s three-tub sink; they’ll be fitted with plexiglass to make the inside of the sink visible to the audience. Klauber is making sure the sonar and laser fit in the housing, Iacobucci works on the membrane that will cover Phil’s buttons, and Boswell wields an X-Acto knife, trying to make the printed parts fit together. In the next four days, the team will work around the clock.
MILESTONE 7: FINAL PRESENTATION
December 12. The moment of truth is hours away. “I’m terrified, and I’m not presenting,” says Macomber. The Orange Team has reason to worry. Today is the first time the students have been able to test Phil in Kresge, where water is supplied through a garden hose. And there’s a problem. The hot-water pressure is too high, the cold pressure too low. They try running both, hoping it will balance out. But too much pressure builds up in the hose. It takes more power to turn the valve on and off. In sense mode (when a hand is used to set the water level), the progress-bar LED draws so much power that there’s not enough to control the valve. What’s more, the pressure is causing water to spray out of Phil’s nozzle; that interferes with the sonar signal, so the progress bar isn’t even working properly. Dou sets to work rewriting the code to compensate.
Wallace got one hour of sleep the previous night, but in the lobby of Kresge before the presentations, he’s ebullient. “Lots of stuff came together,” he says. “We have lots of happy students. That’s what it’s all about.” He has shaved off his beard—and most of his hair—and is wearing a suit and tie for the occasion. People walk by without recognizing him.
At 6:30, an hour before show time, Dou is still programming and testing. Pappalardo technical staffers consult with the Orange Team about the water situation by phone, and mentor Robin Miller suggests using a splitter to divert some of the water to relieve the pressure in the hose.
The new code is not working, and there’s no time to fully test it anyway. With about 30 minutes to go, the Orange Team decides to go with just the low-pressure cold water; that will make for a slow demo, but it should work. Dou reverts to Phil’s earlier, tested code. As the audience begins filing into Kresge, Yang and a contingent from the Orange Team test Phil with cold water only. It’s working. For now.
Kresge’s 1,226 seats fill with family members, product design professionals, venture capitalists, students, and alumni. “People are clamoring to get in,” says lab instructor Peter Nielsen. “The Kardashians could walk through the crowd right now and no one would notice.” At 7:30, after a clock on the video screen counts down two minutes to the frenetic Pee-wee’s Big Adventure theme song, Wallace strides on stage to announce the Orange Team, the first presenters.
Ruleman kicks off the presentation smoothly, but for several heart-stopping seconds, there’s a blank screen when it’s time for the video. She waits, presses the button again, repeats her line, and suddenly, it works. Wilson talks through the demo as Parell uses his hand to set the desired water level in the sink. But through the sink’s plexiglass window, it’s clear that the water is not rising.
Parell tries to time his presentation of the technical details to coördinate with the beep signaling that Phil has reached the right level. The beep doesn’t come. The water continues flowing, and the sink doesn’t fill. Ruleman, Wilson, and Parell assume it’s slow because of the low pressure and carry on as if all is going according to plan.
As they finish their presentation and move on to questions, Parell realizes that the sink stopper wasn’t put back into the drain after the last-minute test on stage. He replaces it, and the water slowly begins to rise. As Ruleman is answering a question, Phil beeps. She stops dramatically in midsentence to point out that the water has reached the selected level. Parell pipes up that it helps to put in the stopper if you want to fill the sink. The audience laughs, Parell shrugs sheepishly, and the laughter turns into cheering. The demo didn’t go exactly as planned, but ultimately Phil worked.
The lineup of demos that follow is impressive. Red’s ThermAssist, a personal heating and cooling pad for paraplegic wheelchair users. Blue’s Cobolt, a keyless bike lock. Green’s Walker Prime, a walker that folds into a cane for tight spaces and stairs. Yellow’s FreeRim propulsion system gives wheelchair users control in wet conditions and keeps their hands clean.
Purple’s Ascent (a.k.a. Chair Force One) adds a smart power assist to a manual wheelchair. Silver’s Exhale compresses a full-size suitcase load into a carry-on bag. Pink closes the night with Helmet Hub, the vending machine for bicycle helmets. The presentations are so polished that no one would guess several teams didn’t get their prototypes working until the last 24 hours.
The final question in the Pink Team’s Q&A is planted, a ploy to get Wallace on stage so the students can thank him. “For most of us, 2.009 is probably the most memorable class we’ll take at MIT,” says Dani Hicks ’12. The students (and most of the audience) stand to cheer, shouting, “Wallace! Wallace!” But Wallace quickly shifts the focus back to the students, marveling at what they’ve accomplished. As the audience chants “Ideate. Model. Test!” yellow, pink, and green Angry Birds are launched into the air from the now well-sealed air cannon. And another semester of 2.009 comes to an end.
Epilogue: Although the audience ranked Phil second—a mere 10th of a point behind Helmet Hub—Orange took it in stride. The team was accepted into Start Labs’ C2C, an incubator program that helps inventors turn concepts into companies; the day of the final presentations, the students filed a provisional patent application. Ten of them worked on Phil during IAP, and Macomber developed Four Leaf for his senior thesis. Dahan says Phil will be just the first of many products “Phil Labs” will bring to market. Indeed, by March, the students were developing a product to simplify medication management for patients even as they debated whether to commercialize a revamped Phil for the home bath market.
Alice Dragoon is Technology Review’s senior editor for MIT News.
The inside story of how ChatGPT was built from the people who made it
Exclusive conversations that take us behind the scenes of a cultural phenomenon.
Sam Altman invested $180 million into a company trying to delay death
Can anti-aging breakthroughs add 10 healthy years to the human life span? The CEO of OpenAI is paying to find out.
ChatGPT is about to revolutionize the economy. We need to decide what that looks like.
New large language models will transform many jobs. Whether they will lead to widespread prosperity or not is up to us.
GPT-4 is bigger and better than ChatGPT—but OpenAI won’t say why
We got a first look at the much-anticipated big new language model from OpenAI. But this time how it works is even more deeply under wraps.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.