We deep draw parts for a variety of
applications including electronic,
communication, aerospace, medical and
more. Contact us today to learn how
our precision deep drawing process can
provide you with enhanced production
effciencies and cost savings over
traditional manufacturing methods.Braxton Manufacturing Co., Inc.
Watertown, CT 06795 • Tel: 860-274-
Braxton Manufacturing Co. of California, Inc.
Tustin, CA 92780 • Tel: 714-508-To order, go to:
http://www.braxtonmfg.com/kit
or call 860-274-FREE
Designer’s
Sample Kit
- High accuracy and repeatability, with
tolerances to ±0.00254mm. Surface fnish
to 8 micro inches. Available uniform I.D.
and O.D. Gold Plating or Cladding - Slotting, faring and hole punching for
intricate designs with no additional
handling - Braxton can deep draw most conventional
and exotic metals and alloys, plated or
unplated, to 57mm max length - Diameters can be as small as 0.215mm
O.D. with a wall thickness as little as
0.0127mm. Length to diameter ratios up
to 57:
ISO 9001:
CERTIFIEDVisit our website: http://www.braxtonmfg.com
the pulses. That gives the circuits the
energy-efficiency advantages of analog
circuits with the accuracy and durability
of digital devices.
“The size of the chip is reduced by
half, and power consumption is one-
third what traditional digital chips need,”
Raychowdhury explains. “We usedseveral techniques in logic and memory
design that cut power consumption to
the milliwatt range while still meeting
performance targets.”
With each pulse-width representing a
different value, the system is slower than
digital or analog devices, but Raychow-
dhury says the speed is sufficient for thesmall robots. “For these control systems,
we don’t need circuits that operate at sev-
eral gigahertz because the devices don’t
move that quickly,” he notes. “We sacri-
fice a little performance to get extreme
power efficiencies. Even if the chip oper-
ates at 10 or 100 megahertz, that will be
enough for our applications.”
The 65-nanometer CMOS chips
accommodate both kinds of learn-
ing appropriate for a robot. It can be
programmed to follow model-based
algorithms or learn from its environ-
ment using a reinforcement algorithm
encourages better performance over
time. It’s much like a child who learns to
walk by bumping into things.“You start the chip with a predeter-
mined set of weights in the neural net-
work so the robot it controls starts from
a good place and not crash immediately
or give erroneous information,” Ray-
chowdhury explains. “When you put
it in a new location, the environment
will have some structures the chip will
recognize and some it will have to learn.
It then make decisions and will gauge
the effectiveness of each decision to
improve its ability to move.”
Communication between the robots
let them collaborate to seek a target. “In
a collaborative environment, the robot
not only needs to understand what it
is doing, but also what others in the
same group are doing,” Raychowdhury
says. “They work to maximize the totalNewsA Georgia Tech researcher places a robotic
car controlled into a test arena where it will
learn and collaborate with another robot.
(Photo: Allison Carter, Georgia Tech)12 MAY 2019 MACHINE DESIGN