My design classes at MIT were in the Department of Mechanical Engineering, and were largely project based. We focused on things like ergonomics, design for manufacturability, and - believe it or not - selling your idea to the client. My design classes in the Communication Department at Stanford were deeply theoretical, and covered things like Fitt's Law and Grice's maxims. I also studied things like production values and cinematography, but that was in the Computer Science Department. Go figure.

My industrial engineering and journalism background have come in handy more than once during product design exercises. The assignment for my last graduate design class at MIT (I was an undergrad, but had run out of undergrad classes) was to design a handheld medical information device for use by doctors and hospital staff. Rather than interview nurses, which is what the rest of the class did, I hung out in an emergency room for an afternoon, and followed one patient through the triage and treatment process (with his permission, of course).

I discovered that the traffic patterns and information staging processes in the emergency room were a disaster waiting to happen. Patient information was kept on identical clipboards, which were supposed to stay in each exam bay but were in fact usually accidentally carried by the examining physician to their next assignment. Lab results were accessed by using a single centralized phone - residents awaiting important diagnostic tests made repeated calls down to the lab until the information was ready, traveling across the floor each time, often dialing without removing contaminated gloves. The same phone was used to call upstairs to the head nurses on the units, to see if rooms were available for check ins.

Patients were also repeatedly asked to recount their presenting problem. This at first seemed an indication of inefficiency, but after looking over a few shoulders I figured out the ploy - the examining residents knew the problem from the triage nurse's notes. Speaking to the patient offered an opportunity to assess lucidity and responsiveness, and to make personal contact.

I designed a "digital clipboard," in matte white with a friendly green stripe around the edge. The device used a clamshell design (two pieces of vacuum formed plastic, screwed together with the electronic guts inside). Input was by menus and handwriting on a touch screen, with the option of storing the handwriting as a graphic, or interpreting it as text input ("sometimes," one of the nurses said, "doctors like to draw pictures"). To distinguish between handwriting to be stored graphically, only in a patient's chart, and handwriting on the same screen to be interpreted as text (for cross-checking with a medication menu, for example), I used a modern-day cartouche system, which allowed users to choose if their text was to be interpreted without interrupting their writing flow.

The device was gasketed to allow it to be fluid resistant - if necessary, the whole thing could be put in a clear plastic bag, and the input done with a tongue depressor. The battery for the device (and hence most of the weight) was in a hemispherical bulge on the upper back - the device could be recharged by induction (no open ports to collect bacteria). The bulge plus the lower edge of the device created a point-and-line support for use on desktop or bedside, and allowed the user to hold the bulge in the non-dominant hand and write with the device resting on that forearm. Woody Flowers (the professor for the class) said the whole thing was "wonderfully stable."

In addition to storing and displaying the patients' chart in electronic format, the device incorporated a beeper (so the lab could notify the resident when results had come in), an IR communication system (since RF would be unworkable in a hospital), and a bar code reader that doubled as the input stylus. The bar code reader was to be used to swipe a code on a patient's hospital bracelet (already printed out by laser printer during triage). This would pop up the patient's chart (which could also be accessed via menu), and would give each person who examined the patient an excuse to take their hand while making an introduction.

I suggested that bar codes also be put on medical supplies, so that bottles and IV bags could be swiped while the patient's chart was open, allowing automatic cross-checking of medications and doctors' orders.

My design solution went over well. Two of the class' project reviewers - principals at Design Continuum and IDEO - suggested I apply for jobs with their firms (I went to grad school instead). A third reviewer, a nurse, said the design would save lives.

Epilogue: While the IR communication system is now familiar to Palm users, the barcode for medication tracking is also being implemented. As far as I know, nobody has adopted the cartouche strategy.

Epilogue Two: The Palm has now been replace by the iPhone, and QR codes have superceeded barcodes. AFAIK, there's still nobody using the cartouche system.