‹#›
U.S.
Department of Transportation
Quality
Assurance: Key Insights
§Collect redundant data from multiple sensors
□Data
can be combined so that false positives are filtered out
§Use standard industry reference files when possible
□Reduces
erroneous information
§Data quality is highly industry dependent
□Choice
between high‐output, real‐time data,
and scrubbed, pseudo real‐time data
□Greater
the overall veracity required,
the more process‐intensive it is to enforce that
veracity, and the higher the delay in
disseminating the data
□Typically,
fast, general data‐quality analysis
for
real‐time
systems, and thoroughly scrubbed
and sanitized data for historical and post
real‐time analysis and
display
Timely data is worth
nothing, if
it is inaccurate