---

Ford and NASA Bring Artificial Intelligence to Cars and Trucks

10 September 1998

Ford and NASA Bring Artificial Intelligence to Cars and Trucks

    DEARBORN, Mich., Sept. 10 -- Ford Motor Company and
NASA's Jet Propulsion Laboratory are bringing artificial intelligence to
trucks and cars.
    A new neural network computer chip that mimics the human mind promises to
reduce vehicle emissions and improve fuel economy by monitoring fuel
combustion.  The neural network chip, designed by computer scientists at
NASA's Jet Propulsion Laboratory (JPL) and licensed by Ford Motor Company, has
the potential to augment current vehicle on-board diagnostic systems.  The
chip may diagnose and control emissions behavior on a continuous basis.  JPL
and Ford scientists contend that the chip represents the most significant
change in the way computing is done on vehicles since computers were
introduced into automobiles in the mid-1970's.
    "For misfire diagnostics, you have to observe and diagnose over one-
billion engine firing events in a car's lifetime.  Additionally, your
diagnostic error rate has to be extremely small, less than one in a million,
in order to avoid sending false alarm signals to the driver," said. Dr. Ken
Marko of Ford's neural network team.
    "This chip will 'learn' diagnostic tasks bypassing the need to develop
conventional software," said Marko.  "It will perform better than the
conventional software and can be used in large production volumes with
standard microprocessors."  Ford expects to be able to offer this new
technology to consumers at no additional cost.
    The chip is derived from Ford's computerized neural network systems, which
are modeled after human learning processes and allow simultaneous processing
of multiple tasks.  This feature allows us to monitor emissions on a continual
basis to alert the driver when excessive emissions may occur due to misfires.
    The neural systems are modeled after the structure of animal nervous
systems.  These systems are comprised of neurons that facilitate inputs being
processed in parallel.  Previously, computers were based on the standard
serial processor.  Under this system, computers could only process information
one input at a time.  By dynamically gauging fuel distribution, this new chip
may provide optimum fuel economy and emission levels, and serve as a
preventive measure for engine misfiring.
    Besides helping to meet the Clean Air Act standards, the neural chip has
the potential to increase fuel economy by improving fuel distribution.  The
chip also promises to improve driveability.  By significantly reducing the
chance of misfiring, it decreases emission levels and virtually eliminates
false alarms the vehicle may signal.