James McLurkin has found the perfect parking space for his car. It's convenient, it's easy to find, and it's incredibly close to home. But trust me, you don't want any part of it. McLurkin's car is--how shall I put this?--something of an off-road vehicle. See, the place McLurkin wants to drive his car is your colon. You may never have thought of your intestinal tract as a racetrack, but McLurkin has, and the unusual vehicle he's developed could forever change the face--or whatever--of modern proctology.
Proctological science is something of a mystery to most people, and they'd like to keep it that way. Whatever reason a medical professional would have for choosing to enter the field ("Billy was always fascinated by colonic polyps!"), it certainly can't be in order to become popular. During the average proctological exam, there are few opportunities for hearty conversation, fewer still for an exchange of the latest jokes, and eye contact is out of the question.
Then there are the tools of the proctologist's trade. It tells you a lot about the medical experience you're about to undergo when the least threatening item in your doctor's office is a latex opera glove. Perhaps the most feared of all procto-gadgets are the proctoscope and sigmoidoscope--diagnostic tools the shape and size of commuter trains that even proctologists admit cause "a little discomfort." Those who have had this kind of exam feel more strongly about it, pointing out that while the folks at Amtrak may urge travelers to take the Metroliner, they don't mean in suppository form.
Not even a proctologist will suggest that an intestinal exploration is an experience to look forward to. ("Troy Aikman, you've just won the Super Bowl! What are you gonna do next?" "Undergo extensive sigmoidoscopy!") Like their patients, these doctors have long wondered whether there might be a way to make the entire ordeal at least a little less unpleasant. Is it possible to develop a proctoscope that feels a bit less like a periscope? A colon exam that's less grueling than a bar exam? James McLurkin thinks he's found a way.
McLurkin is an MIT senior majoring in electrical engineering, minoring in mechanical engineering, and specializing in a field that requires extensive skills in both areas: robotics.
"Even before I started school, I was intrigued with the idea of robots," he says. "Building mobile machines that could perform tasks autonomously or semiautonomously seemed like the ultimate engineering challenge. What held special appeal for me was microrobotics."
The popular image of robots is a somewhat limited one. The average robot is a six-foot bipedal device that, if you believe the writers of The Day the Earth Stood Still, acts more or less like an ordinary human being, except that it speaks fluent Esperanto. By the time the television series Lost in Space rolled around, the archetypal robot had warmed up some, even mastering English--provided, of course, the English was some variation on the words danger and Will Robinson. Ultimately it was The Jetsons that offered the most appealing image of the future of robots, as long as you were willing to accept that the entire robotics industry would devote itself to building an artificial human that looks and sounds exactly like Shirley Booth.
But robots don't have to be so anthropomorphic. "When I first got to MIT in 1991," McLurkin says, "some of the folks here were already working with small robots about the size and shape of a toaster." The machines, called toaster-bots, were miniature versions of robots that might someday do grunt work such as the terraforming that would be needed to build a lunar or Martian base. Leaving aside the thorny question of whether it would be a good idea for NASA to send kitchen appliances as humanity's emissaries to the stars, McLurkin saw another, more immediate problem with the toasterbots: they were too big.
"Toaster-size doesn't sound like much when you're talking about robots," he says, "but the bigger the machines, the bigger you have to make the lab or room where you test them. If you downsize your experimental robots, you can test them on something as small as a tabletop; then, when they're perfected, you can scale them up to whatever size you want." Working with tiny custom circuit boards and gearboxes, McLurkin started building a fleet of seven palm-size robots he called the Dwarves, which could cavort in a three-by-three-foot sandbox.
Outer space isn't the only frontier for robots, though--there's also inner space. For years engineers have been contemplating the idea of miniature medical robots, perhaps floating through your veins and arteries prowling for blood clots. The twisted course of the bowels has also attracted its share of armchair explorers. In 1992 Joel Burdick, a mechanical engineer at Caltech, and Warren Grundfest, a physician at Cedars-Sinai Medical Center in Los Angeles, developed a 17-foot, 110-pound intestinal snake intended to slither comfortably and autonomously through a patient's intestines in order to diagnose disorders. The autonomous part the inventors mastered well enough, but the comfortable part was another matter, unless they expected a barosaur to be coming by for its annual checkup. Of course Burdick and Grundfest intended the device (called "Snakey") merely as a demonstration model, and they planned to downsize it dramatically.
McLurkin, meanwhile, got into bowel navigation by accident. In 1992 he created a robot--dubbed Goliath--that was even smaller than the Dwarves. Built with no specific mission in mind, Goliath measured just 1.1 inches by 1.2 inches by 1 inch high. To power it, McLurkin used two miniature motors originally designed to make personal pagers vibrate. One afternoon, when McLurkin was off at class and the completed Goliath was sitting unattended on his workbench, his adviser, Anita Flynn, was giving a surgeon a tour of the lab. When he spotted Goliath, his eyes lit up.
"He immediately turned to Anita," McLurkin says, "and said, 'Hey, this thing might be perfect for the kind of procedures I perform.' "
Much of what an intestinal surgeon does, he explained, involves diagnostic or repair work. Polyps have to be biopsied; ulcers have to be spotted and evaluated; other kinds of lesions have to be treated. If a robot like Goliath was equipped with manipulating arms and calipers, a miniature camera, and, not insignificantly, a light, it could perform many of those chores. The robot would trail a tether of various wires and hoses, but still, it would be far more comfortable for a patient than a high- caliber sigmoidoscope.
As Flynn listened, she was slowly persuaded. Sure, McLurkin had his sights set on the cosmos, but would this be so different? All he'd really be doing would be swapping the dark side of one moon for another. When Flynn proposed the change of venue to McLurkin, he was as excited as she. They began divvying up the work that would be needed to make this singular dream come true. Flynn and engineering student Dean Franck would further miniaturize Goliath's micromotors; MIT senior Art Shectman would develop the lights, sensors, steering mechanisms, and other onboard equipment that would allow the robot to move; and McLurkin would figure out how to integrate these and other components into a working model that could survive in an environment that is not exactly the garden spot of the body.
To familiarize himself with the terrain his machine would be treading, McLurkin sat in on an endoscopic examination of a patient who apparently had an extremely high embarrassment threshold. During this procedure, a camera attached to a flexible probe is inserted into the colon ("I'm ready for my close-up, Mr. DeMille!"), and the images it picks up appear on a television screen. For the untutored observer, the experience is something less than prime-time programming, and McLurkin, not surprisingly, soon wished he had opted for Cheers.
"In case I ever had any doubts," says McLurkin, "I learned pretty quickly that I am not cut out for medicine. As soon as the procedure started, all the blood just fled from my head."
Before his own screen went blank, however, McLurkin was able to catch a few glimpses of the television screen, and what he saw was not encouraging. "The large intestine turns out to be an awful environment for a robot. It's wet, it's slimy, and you can't push against the wall to get traction because it will just stretch underneath you. What's more, even if you could get an ordinary machine to operate under these conditions, this machine has to be sterilizable. Any one of those things is difficult to accomplish. Together, they're nearly impossible."
Despite these problems, in the two years since McLurkin's introduction to the duodenal diamond lane, he and his colleagues have gone a long way toward designing a vehicle capable of navigating it. The machine he's blueprinted and partially built has been dubbed Cleo, and if Goliath was an intestinal Edsel, Cleo is a Cadillac.
"Like military tanks," says McLurkin, "Cleo moves about on two treads and turns with the aid of what is known as differential steering: If both treads move forward or backward, the tank does, too. If the treads move in opposite directions, the tank turns left or right, depending on which tread is moving in which direction." Around the robot is an array of sensors: four to detect visible light, four to detect infrared, one to detect tilt, and others to feel for obstructions. Extending from the front is a claw that allows the robot to grasp and carry objects. The machine carries an onboard battery that is able to run all this equipment. In its trailing umbilicus are an air hose, a vacuum hose, a video cable, and a power line for a camera and a floodlight.
As McLurkin and his colleagues envision things, a workday for the robot would begin when Cleo entered the body through its least desirable toll plaza and, with high beams on and camera whirring, drove to the site in need of roadwork. Once the intestinal pothole was spotted by the camera and the image was beamed back to a television screen on the outside, the surgeon would use the air hose to slightly inflate that part of the intestine so the robot would have room to move. The surgical arm would then perform whatever work was necessary, while the vacuum hose would clean up afterward. The robot could simply drive back out without leaving so much as a traffic cone or a thermos of coffee behind.
At the moment, the lights, camera, and hoses are still in the R&D; phase, but the basic hardware of the car has been up and running for months. On a table in McLurkin's MIT lab is a hollowed-out model of a large intestine that serves as a racetrack for Cleo. With the aid of a joystick, McLurkin puts the car through its sinuous paces, hoping to learn more about its steering abilities and how they can be improved to accommodate even the sharpest intestinal turns.
"Cleo is going to have to do a lot of complicated driving to get where it's going," he says. "At the moment, I'm trying to make sure its sensors and steering are up to the job."
Once Cleo masters the maze and is fully outfitted with its additional equipment, it still won't be anywhere near ready to get its kicks on your Route 66. The current car, while impressively small, is still larger than McLurkin would like. Even an inch-long suppository can be intrusive when it's the kind of suppository that requires a motor vehicles registration sticker. And so before Cleo can begin its less-than-fantastic voyage, McLurkin, Flynn, Shectman, and Franck want to get it down to thumb size.
In addition to the size and comfort problem, there is also a safety problem. A robot that is going where Cleo is designed to go will have to be moisture-proofed, not just to keep it clean but to spare the patient the ultimate in joy-buzzer experiences from the onboard battery. Sealing the entire machine in a membrane, however, while still allowing it to drive and cut and grasp is a complex problem, and the MIT team is nowhere near solving it. Happily, the other main danger Cleo presents--the danger of disappearing when it's off on one of its missions--has already been taken care of, since the robot will be connected to the outside world by its leash of hoses and cables. This should reassure doctors and patients that Cleo will not just wander off the reservation in the middle of a job and begin beaming back images of the kidneys, the pancreas, or the Pyrenees. "It goes without saying," McLurkin concedes, "that you wouldn't want a robot lost in the large intestine."
Of course, before Cleo can get lost in a large intestine, it will have to get into a large intestine. While the creators of Snakey successfully downsized their device and introduced it into the intestine of a luckless pig this year, McLurkin and his colleagues still have a welter of engineering puzzles to solve before they'll be ready to take such a plunge. Shouldn't Cleo, of all cars, have an emissions control device? What about a moon roof? What about an I'd Really Rather Be Sailing bumper sticker? Until these questions are answered, nobody at MIT would want to try out the device in a real live colon--human or otherwise--and the patients who would be asked to provide the organic interstates would probably agree. After all, we've all had an intestinal bug before, but who ever thought it would be a Volkswagen?