REDMOND, Wash., Feb. 19, 2001 — Lyndsay Williams blends into the crowds of shoppers and students as she rides her bicycle through the winding streets of Cambridge, England. But the research she sometimes does while pedaling to work at Microsofts Cambridge Research Lab may one day make Pocket PCs and cell phones smarter.
At Microsoft’s Cambridge Research Lab, researcher Lyndsay Williams works on technology that will make mobile devices smarter and able to respond automatically to users’ needs.
Half a world away, at his work bench on Microsofts Redmond campus, Mike Sinclair is making personal computing easier and potentially less expensive without writing or revising a single line of software code.
Williams and Sinclair are part of a small group of Microsoft Research (MSR) employees who are doing something almost unheard of at many software companies: They are improving and expanding the horizons of computer hardware.
In addition to helping create prototype devices to run new Microsoft software, these researchers are developing new ways to control and interact with PCs and scores of other devices. If their current research and areas of interest bear fruit, heres a taste of what computer users can expect to see in the future: inexpensive computer displays as small as a fingernail, pens that function like mini-computers and handheld devices that know when to hold your calls and switch themselves on and off.
“People are pretty complacent with their mouse and keyboard,”
said Sinclair, a senior researcher within MSRs Hardware Devices group.
“We need to push the limits of computer hardware and find new and better ways to control and interact with the PC and other devices. Thats what we are here to do.”
Microsoft Saw Potential of Mouse Early
Although Microsoft designs and markets a growing selection of keyboards and mice, the companys high-profile software often obscures its role in hardware development. A prime example is Microsofts early championing of the computer mouse.
Although Apple Computer was the forerunner in the commercialization of the mouse and graphical user interface, Microsoft incorporated their use into early versions of Word in the early 1980s, before most computer users had used a mouse. When computer hardware makers balked at the idea of manufacturing mice, Microsoft had them manufactured independently and stockpiled thousands in a warehouse until the new devices popularity grew.
“We were selling mice way before Windows,”
said Charles Simonyi, a 20-year Microsoft employee and one of only 17 who have been named Distinguished Engineers by the company.
“The idea wasnt that we were trying to be a hardware company,”
Simonyi explained.
“The idea was that we would use hardware to enlarge the utility of software. Thats the fundamental distinction.”
The same is true with Microsofts development of the SoftCard, which enabled Apple computers to run Microsoft software, he said.
“It just takes someone with a good eye, who can recognize where computing is going, to champion new technology, whether it is software or hardware,”
said Simonyi, who now works on intentional programming, a programming environment that strives for maximum reuse of software components.
Making Mobile Devices Smarter
Williams work at Microsofts Cambridge lab is an example of this forward-looking approach — though frustration also played a big role when she began her latest endeavor, a project internally code-named SmartMoveX.
“Its an annoyance to have to remember to switch off my mobile phone when Im in meetings and then remember to switch it on when I come out,”
she said, sitting among soldering irons, electronic feedback machines and assorted mechanical parts in her office in the heart of the citys downtown shopping district.
As envisioned, SmartMoveX will track the users location and actions and use this information to control other devices. If a person is sleeping, SmartMoveX will ensure her mobile phone sends calls to voicemail. It will instruct a persons Pocket PC to display a list of household chores as he walks through the door at night.
“The great thing is that all of this will be invisible to users once they set up the notification requirements,”
Williams said.
SmartMoveX can work invisibly because it doesnt rely on large or obtrusive technologies such as video cameras or global positioning monitors. Instead, Williams prototypes use smaller, simpler technologies, such as light and heart-rate sensors and accelerometers. Smaller than a fingernail and weighing less than a gram, these inexpensive micro-machines measure movement and are most commonly used to trigger car airbags.
Williams attaches the accelerometers and other monitors to a computer chip board roughly the size of a pager, along with a tiny controller, a memory chip and a battery. During early research, she wore the prototypes for as much as 22 hours a day for up to five weeks.
“I dont mind having things stuck to me,”
she said with a laugh.
Each day, she downloaded the contents of the memory onto her desktop PC to analyze and synchronize the information with the calendar in Microsoft Outlook. For a while, she built a small radio transmitter into SmartMoveX to automatically feed the information to her PC. It didnt take long for her to notice patterns, such as spikes in her heart rate when she read important email and, not surprisingly, when she was riding her bike to work. The biggest spike came when a blown tire caused her to crash her bike.
Williams plans to build a smaller version of SmartMoveX that fits into other mobile devices. Then, with the help of software, the patterns within the readouts will help trigger controls on the devices such as on/off switches, schedule reminders and automatic voice mail routers. And, Williams is pleased to say, it may even be able to synchronize with her electronic calendar to tell her mobile phone to switch off when she leaves her office and enters a meeting.
“Advances such as SmartMoveX will allow computers to become much more like personal assistants than tools we must constantly monitor,”
said Roger Needham, managing director of the Cambridge lab.
“Once these technologies are perfected, they will be able to check my personal calendar and determine if Im available to chat with someone who is requesting a meeting.”
Maximizing the Potential of Current Technologies
If you look around the work areas of MSRs hardware researchers, you wont see lots of newfangled gizmos and gadgets. Youll mostly see bits of tried-and-true technology with untapped potential.
“We look for technologies that are not being used to their full potential, find ways to maximize this potential and then work with product groups to find a market for our ideas,”
Sinclair explained.
Sinclair and fellow MSR researcher Gary Starkweather spotted more untapped potential when they began working with tiny electronic mechanical systems. Now they are combining these systems, called MEMS, with microscopic mirrors and lasers to create computer displays of virtually any size, even as small as a fingernail. Starkweather, who invented the laser printer while working at Xerox Palo Alto Research Center (PARC), expects that improved displays, along with new methods for storing and retrieving printed information, will increase worker productivity and make people less dependent on paper.
“While I do not envision the paperless office anytime soon, its not unreasonable to expect that improved displays will reduce the use of paper,”
Starkweather said.
Cost is another incentive: These displays would carry a much smaller price tag than those built using other technologies, Sinclair said. Similarly, he hopes to integrate the technology used in inexpensive flatbed scanners into digital cameras. The goal is to create much less expensive panoramic cameras.
Finding easier ways to interact with the desktop PC is another focus for MSRs hardware researchers. They are developing new ways to control keyboards and researching ways to allow users to control their computers with facial expressions and motion detectors. The detectors may one day instruct computers to increase the size of type on the screen when the user moves away from the monitor.
Williams, meanwhile, is developing pens that convert handwritten messages to computerized text. She has developed improvements that track the writers hand movements more accurately than previous technologies. Her latest prototype also transmits these movements automatically to a PC, where they are converted into text.
Hardware Researchers Work Alone and With Product Groups
Many times, MSRs hardware researchers have no final product in mind when they begin working on a new project.
“If we were to limit ourselves to an exact embodiment of an end product,”
Sinclair explained,
“that would be an undue burden, and it would limit the scope of our research and the potential end products.”
The researchers also work in tandem with Microsoft product teams to build prototype devices and retrofit current ones to test new software. Sinclair is known as the MacGyver of Microsoft, after the resourceful character on the television show of the same name.
“He can cobble anything together with bubble gum and a toothpick,”
joked Ken Hinckley, a researcher in MSRs Adaptive Systems and Interactions group.
Hinckley has worked with Sinclair to develop new mouse and keyboard controls. Recently, he needed help adding sensors to a handheld computer.
“They do enough to get you started, and you are off to the races,”
he said.
This kind of collaboration helps software engineers see software and hardware as more of a continuum, rather than two separate worlds, said Hinckley, who often meets with the hardware researchers to pick their brains.
“They really open your eyes to the kind of hardware we will be using in the future,”
Hinckley said.
“Thats important when you are trying to create the software that will run this hardware.”
