As a matter of introduction I am a twenty-year law enforcement officer who
for the last fifteen years has been instructing law enforcement driver
training and investigating accidents. I also, at the age of forty something,
am about to complete my engineering degree. I have combined my education and
vocation into a study on law enforcement drivers. This study is being
conducted under the direction of the College of Mechanical Engineering at
Florida Atlantic University, with the cooperation of my agency, the Palm
Beach County Sheriff's Office.
Context and Background
The Palm Beach County Sheriff's Office is headquartered in West Palm Beach,
Florida. It has more than 2500 employees, of which approximately 1000 drive
sheriff's office motor vehicles (predominantly marked police cars). This
fleet operates more than 30,000 vehicle miles per day. The fleet is involved
in more than one collision every day. Vehicle collisions are the sheriff's
largest financial and personal injury exposure.
This research project has taken the following form. Once each year, all law
enforcement employees attend a training week. During this week each officer
is exposed to a smattering of relevant law enforcement training, legal
updates, firearms, etc. Four hours of this year's training week is dedicated
to the driver study. The first goal for this study is to document the
current state of the fleet driver. Each driver is measured in many different
ways.
First, each driver drives over a closed course in an instrumented car. The
instrumentation is roughly equivalent to a Flight Data Recorder. An onboard
PC equipped with a data acquisition card measures the driver control input
across 3 channels.
a) Throttle Position
b) Brake Line Hydraulic Pressure
c) Steering Wheel Movement
In addition to these control inputs, the vehicle's response is measured by
two perpendicular accelerometers. With these five sensors the driver's
inputs and vehicle response are documented as a data file on the PC.
Before each run the driver is given the motivation that he is to drive at a
maximum effort (get through the course as quickly as possible). A typical
run time of 2 minutes generates 30,000 data points. This hardware and
original software is the engineering part of the study.
In addition to the hard performance data, all students also take a
multiple-choice survey. The survey's intent was to document each student's
background, training, and experience. The results of this survey are
electronically graded. This electronic grading allows the survey responses
to be easily analyzed through automated techniques.
At this point we are about one year into the study. The first six months
were spent in the preparation of the program. Data collection began Jan. 1,
1998.
Preliminary Results
The data acquisition portion of the project is documenting extraordinary
variability in driving performance. Anti-lock Brake System usage, steering
rates, brake profiles, foot movement time· ad nausium. I would be happy to
elaborate on these topics if some one is interested. They are tangential to
this request.
The survey has also generated an interesting profile of the Law Enforcement
driver.
The answers to two questions on the survey have generated a great deal of
interest. We asked the following question of our law enforcement drivers.
1) Most vehicle collisions could be avoided if an involved driver changed a
decision prior to the collision.
Astonishingly, less than half of our drivers completely agree with this
statement.
2) Out of the total number of vehicle collisions, what amount are
preventable by a change in an involved driver's decision? Again to our
amazement, one in four of our drivers believe that only half of all
collisions can be prevented by a driver's action.
These glimpses into our fleet driver's ideas on collisions are disturbing.
Clearly, if we are to improve our accident rate we must first instill in our
drivers the belief that they have the ability to influence that outcome.
Enter Error Management, Human Factor training.
Specific Request for Help
What began as an engineering project has moved away from that discipline.
This note is an explicit solicitation is for a research partner for this
project. We are in need someone whom is familiar with non-engineering
research techniques such as survey writing, human performance, and
analytical methods.
Our training is committed to the long term. A follow up program of continued
training, continued testing and correlation to accident statistics is
foreseen. The goal of course is to determine success of the program.
Close
Applying EM HF training to Law Enforcement Drivers places us in a unique
position. While this type of training is widely accepted in other contexts,
those error rates approach statistical insignificance. We on the other hand
have plenty of well-documented error. If HF training is successful in
mitigating our error rate, Law Enforcement drivers may be the group that can
be cited to show that CRM trainers are accomplishing something rather than
merely raising the warm and fuzzy quotient.
Please feel free to forward this request to a third party that may be of
assistance. I can be reached in any of the following ways.
Work phone- 561 688 3687
Home phone 561 965 1235
Or via e-mail at bwright_at_tarorigin.com
Thank you for your efforts on my behalf.
Bill Wright
bwright_at_tarorigin.com
http://www.tarorigin.com
561 965 1235