麻豆传媒

I鈥橫 ON Easy Street. However, from where I鈥檓 sitting, it looks anything but.
Easy Street is a lab in a renovated mill in Manchester, New Hampshire. Only you
won鈥檛 find lasers and test tubes in this open-plan, single-storey room, only a
very unusual assault course. There are stairs everywhere, lots of ramps, kerbs
and even a long thin sandpit.

This strange lab houses Project Fred鈥攁 secret endeavour started six
years ago by Deka, an engineering and development company. More than 60 of
Deka鈥檚 staff now work full-time on Fred. The project has attracted a commercial
partner in Johnson & Johnson and consumed more than $100 million in
capital. And I鈥檓 sitting in the result鈥攖he world鈥檚 most sophisticated
wheelchair.

Deka refers to its invention鈥攖he Independence 3000 IBOT鈥攏ot as a
chair, but as an all-terrain transporter. At first glance the IBOT looks similar
to a typical electric wheelchair. The main difference is that instead of two
very large rear wheels, it has four small powered wheels, two on each side, plus
two small castor wheels, one below each of the passenger鈥檚 feet. But this isn鈥檛
simply a piece of rolling furniture: it鈥檚 a sophisticated vehicle with on-board
intelligence and powerful four-wheel drive that would be the envy of Detroit. It
even has an innate sense of balance. Most importantly for disabled people, the
IBOT is going to equal the stakes. They鈥檒l be able to travel faster, negotiate
irregular surfaces with ease, and if they want to reach the top shelf, they can.
The IBOT allows its passenger to feel six feet tall, literally.

It鈥檚 time for project manager Lucas Merrow to demonstrate the IBOT鈥檚 powers.
He hops into the seat and enters a few quick keystrokes on the control pad built
into the armrest. This is to program the IBOT to respond for his body weight and
height. Motors and gears hum into action, driving the transporter gently over a
kerb.

But Merrow doesn鈥檛 tilt dangerously backward as he would in a standard
wheelchair. The passenger seat tilts, keeping him perfectly level. Next, the
IBOT breezes up an 18-degree ramp from street level to an elevated platform.
鈥淲ay steeper than allowed by US law,鈥� he says.

Using a joystick alongside the keypad, Merrow turns the transporter in a
circle over the rough cobblestones on the platform. It spins around on the spot.
In a normal wheelchair 鈥渁 disabled person would need a lot of upper-body
strength to do that鈥�, says Merrow. Then he pushes the joystick forward and
plunges into the sand trough. Amazingly, he and the IBOT sail through the
20-centimetre deep sand like a sailboat on water. 鈥淲e had a disabled man in his
regular chair try to do that,鈥� says Merrow. 鈥淗e had tremendous muscular
development in his upper body and he couldn鈥檛 get much farther than a metre.鈥�
Next, Merrow backs up to the bottom of a staircase and taps a few keys. This is
where the IBOT really starts to show its skills. As Merrow grabs the stair鈥檚
handrail beside him, the front-drive wheels rise up and revolve over the ones
behind. As these wheels meet and settle on the tread of the first step, the rear
pair repeat the action, pulling the chair upwards and backwards. With Merrow
gripping the handrail for control, the wheels revolve backwards over each other
in a kind of mechanical hand-over-hand鈥攍ifting chair and passenger up the
stairs. To come back down, Merrow simply reverses the action and the IBOT rolls
back down: the rear-drive wheels rise up, over and down past the front-drive
wheels to settle on the step below and so on.

But Merrow has saved the most dramatic feat for last. Back on level ground,
he suddenly rears up and throws himself backwards in the chair with as much
force as he can muster. The chair begins to tip backwards and I lunge to grab
him.

Before I can reach him, the IBOT rises to the occasion. A motor under the
seat lifts the transporter鈥檚 two front-drive wheels off the ground鈥攋ust as
they did when Merrow started to climb the stairs. But the motor stops when the
wheels are directly above the other pair. So now, instead of the drive wheels
being side by side, one pair is atop the other
(see Diagram, p 30).

Within seconds, I鈥檓 looking Merrow in the eye. Now that the transporter is
raised up on two wheels, he is sitting at standing height. The IBOT is balancing
on two wheels the way that most people balance on two feet.

鈥淚t doesn鈥檛 take that much force to raise the IBOT to two wheels,鈥� says
Merrow. 鈥淵ou can do it with a shrug, a push off a solid object, or by tapping
the keypad. If you鈥檙e on a smooth, level surface, you can ride around on two
wheels as easily as if you were down on all four.鈥�

As Merrow stirs and shifts in the seat, the two wheels on the floor move
constantly, inching back and forth to keep him upright鈥攖he mechanical
equivalent of a person unconsciously shifting body weight from foot to foot
while standing still. Suddenly Merrow throws himself backward again, but the
transporter races back to stay under him. He starts to fall forward out of the
chair and it scoots forwards to catch him.

He taps a few keys to settle the IBOT back down on four wheels. 鈥淵ou try it,鈥�
he says. It takes me a few tries鈥攁bandoning your physical safety to the
intelligence of an inanimate object takes some doing. Once I鈥檓 balanced aloft in
the IBOT, Merrow holds up his hands, palms toward me. 鈥淧ush,鈥� he commands.

We lock fingers. I push. The machine beneath me doesn鈥檛 roll back鈥攊t
stands its ground. 鈥淗arder,鈥� Merrow urges. I push harder, as hard as I can, and
after a few moments Merrow staggers backward under the force. 鈥淯nder any but the
most extreme force, the IBOT won鈥檛 let you fall,鈥� he says. The secret lies not
only in the durability of its motors and gears, but also in the host of sensors,
solid-state gyroscopes and three Pentium-class processors mounted beneath the
seat. The sensors 鈥渇eel鈥� things, for example when the wheels have run up against
the riser of the next stair. The gyroscopes determine balance and the
processors make the million tiny decisions that allow the IBOT to get
around.

If you look at the chair and its rider from the side, you can imagine a
pyramid: the apex is the driver鈥檚 head and the front and rear of the chair鈥檚
base form two of the bottom edges of the pyramid. The job of the processors and
the gyroscopes is to make sure that the pyramid鈥檚 centre line remains parallel
to the force of gravity鈥攅ven when the chair鈥檚 chassis tips.

It was back in 1991 that Dean Kamen, Deka鈥檚 founder, first thought of this
remarkable wheelchair. While at a shopping mall, he watched a strapping young
man in a wheelchair try repeatedly to heft himself up and over a kerb. If we can
put a man on the Moon, thought Kamen, why can鈥檛 we get a man in a wheelchair
over a kerb? Kamen wasn鈥檛 one to let the problem go. At college, he鈥檇 patented
the first automatic portable infusion pump for dispensing drugs. At 25, he
founded a medical device firm, AutoSyringe.

Kamen began prototyping a vast array of ideas to solve the wheelchair
problem. 鈥淲alking machines, robotic legs, things you鈥檇 strap onto yourself,
things that were unstable and unreliable,鈥� Merrow says. 鈥淭hat went on for three
or four years.鈥�

The turning point came when Kamen had a near-accident of his own. Getting out
of the shower, he slipped on the wet floor and windmilled his arms to catch his
balance. 鈥淚t dawned on him that in order to have a machine that could do what
he鈥檇 just done, which was to lose his balance and then recover,鈥� says Merrow,
鈥測ou had to have a machine with a sense of balance in the first place.鈥�

So Kamen went back to his lab and built a platform with two legs, with the
whole thing mounted on wheels. It had a bunch of circuit boards with wires
hanging off, some bicycle chains and a pair of motors from an old sewing
machine, Merrow says. 鈥淚t was rickety. It shook and shuddered. But it worked.鈥�
At that point, Kamen assigned a few of his engineers to take things farther.
Someone on the development team realised that they couldn鈥檛 climb kerbs and
stairs using one drive wheel on each side of the chair. So they put two on each
side, added a pair of smaller wheels and slid a desktop PC under the seat.
鈥淭hat鈥檚 the machine we took to Johnson & Johnson in 1995,鈥� Merrow says.

Kamen knew that Deka needed a manufacturing and marketing partner. But they
didn鈥檛 know who. 鈥淲e thought, 鈥榯his is a vehicle鈥�, so we talked to the car
companies,鈥� Merrow recalls. 鈥淏ut the IBOT has orders of magnitude more software,
processing power and back-up systems than a car has. To a car company, the IBOT
looks more like a computer. But to a computer company, it looks more like a
vehicle.鈥� They also talked to defence contractors, who were accustomed to
making complex combinations of hardware and software. 鈥淏ut they were never
comfortable about how their expertise [would] fit the project,鈥� he says. As part
of the search, Kamen contacted Robert Gussin, then Johnson & Johnson鈥檚 chief
technology officer. Gussin saw the transporter鈥檚 prototype and became a convert.
However, his bosses were a tougher nut to crack. 鈥淛ohnson & Johnson had once
owned what is now the world鈥檚 largest wheelchair company,鈥� Merrow says. 鈥淭hey
decided that they didn鈥檛 like that business and they sold it. But Gussin kept
pressing them, telling them that this is nothing like a regular wheelchair, it鈥檚
the kind of thing that can turn an industry on its head, and the kind of thing
that the company should be doing.鈥�

Asking difficult questions

Eventually, Gussin鈥檚 persistence paid off. Johnson & Johnson licensed the
IBOT鈥檚 design and technology from Deka and set up a small company called
Independence Technology to make and sell it. For the next year, an IBOT project
team travelled the US convening more than 30 focus groups of up to 20 wheelchair
users each. 鈥淲e asked them what they like about the chair they use now, what
they wish was different,鈥� Merrow says. 鈥淚f you could have anything you wanted,
what would it be? Which wish would come second? What kind of trade-offs would
you be willing to make?鈥�

Through the discussions and more than a thousand questionnaires, the team
discovered an untapped market: more than two million wheelchair users in the US
alone. 鈥淭hey aren鈥檛 in nursing homes or hospitals, they have more and more
discretionary income to spend, and no one is treating them like customers,鈥�
Merrow says. 鈥淐hair makers deal primarily with insurance carriers, not directly
with users, so they鈥檙e trying to keep insurance companies happy by keeping
prices down. No one has ever bothered to ask disabled people what they want in a
飞丑别别濒肠丑补颈谤.鈥�

The hardest work came next: not just creating the hardware and software to
carry out the tasks the users were calling for, but integrating those parts into
a practical system. 鈥淲e didn鈥檛 have to invent hardware, but we had to work with
vendors to get them to understand our requirements,鈥� Merrow explains. 鈥淔or
example, the company that makes transmissions for the IBOT also makes
transmissions for high-performance cars. Our transmissions needed some similar
characteristics but also some very different ones, such as the ability to spin
gears backwards as well as forwards.鈥�

Two of the three main limitations were power and space. The transporter had
to fit through the narrowest doorways and hold a human frame, yet carry enough
batteries, gears and motors to tote a person weighing as much as 115
kilograms鈥攖he IBOT鈥檚 weight limit鈥攐ver kerbs and up ramps and
stairs. Months went into reshaping and streamlining components, cajoling
suppliers to shave a centimetre here or a few grams there, fitting the entire
thing into a space roughly half a metre square and a third of a metre high.

The third, but most important requirement is safety. 鈥淲e have three
processors so they can vote,鈥� Merrow says. If you鈥檙e going over a kerb, what
command do you send to the motor? How much torque do you apply? 鈥淚f there鈥檚 one
[processor] and it鈥檚 wrong, you鈥檙e in trouble. If there are two and they
disagree, the device comes to a stop.鈥� It鈥檚 the same problem when the chair is
balanced on two wheels. 鈥淚f one processor fails, you fall over. If you have two
and they disagree, the machine freezes鈥攁nd you still fall over. You need
three to validate decisions.鈥�

The IBOT has two nickel-cadmium batteries, each of which can supply enough
power to keep the device moving on its own. Together they will power the
transporter for about 25 kilometres non-stop on a smooth, level surface. 鈥淭hey
were designed to provide a typical user with enough power to get through an
average day,鈥� Merrow says.

There are also three motors. One spins the single axle attached to the two
pairs of drive wheels, enabling the IBOT to climb stairs and balance upright.
The other two motors drive the wheels directly, one motor per pair. If a motor
overheats, the chair automatically slows to a crawl and a yellow warning light
appears on the armrest鈥檚 control panel. If overheating continues, the chair
simply stops. 鈥淚t鈥檚 virtually impossible to burn out a motor on this device,鈥�
Merrow says.

This summer Deka opened a durability testing area the size of a gymnasium.
Two shifts of eight volunteers will each spend eight hours a day using the
chairs non-stop鈥攗p and down stairs and ramps, in and out of vans, through
puddles and sand pits, over rough surfaces, indoors and out. Meanwhile, IBOTs
are being put through a gamut of other tests to check they can withstand the
stresses of everyday life.

Deka and Johnson & Johnson have been feeding their test data to the US
Food and Drug Administration. 鈥淭hey鈥檝e been great,鈥� Merrow says, 鈥渢hey鈥檝e put us
on a fast track for approval.鈥� The IBOT is expected to be certified for
commercial sale as a medical device some time next year. The price tag:
$25 000.

But will it sell at that price, especially in enough volume to recoup the
$100 million that the two companies have sunk into the venture? 鈥淲e can鈥檛
know for sure,鈥� Merrow admits. 鈥淏ut this isn鈥檛 entirely about business. This is
something we couldn鈥檛 not do. No one listens to people with disabilities, but
they want to be treated like anyone else鈥攍ike customers. We鈥檙e going to
give it a shot.鈥�