|
NIST
Project Goal:
Develop metrics, measures, data, and analysis methods
to quantitatively evaluate component technologies and integration strategies
in order to accelerate the development of intelligent behaviors in unmanned
vehicle systems.
DARPA
MARS
Program Mission:
Develop
(learning-based) software technologies required for robust perception-based
autonomy.
|
MARS 
DARPA
Mobile Autonomous
Robot Software
(MARS)
DARPA
MOBILE AUTONOMOUS ROBOT SOFTWARE (MARS)
Metrics and Performance
NIST is developing metrics, measures, data, and analysis methods to
quantitatively evaluate component technologies and integration
strategies in order to accelerate the development of intelligent behaviors
in unmanned vehicle systems.
Our main thrust areas are the following:
1. Task Analysis for Driving
 |
Task
decomposition and path planning methods and algorithms
|
|
Real/Virtual
worlds for on-road driving
|
|
Road
map for autonomous on-road driving technology requirements
|
2. Calibrated Data Collection &
Dissemination
|
Data
collection in various environments and scenarios
|
|
Field
support for autonomous mobility testbeds
|
3. Performance Metrics
|
Ground
Truth and Benchmarking for Evaluation of Sensory Processing
Algorithms
|
|
Technology
evaluation testbeds and performance measures
|
|
Human
Intervention Analysis and Evaluation of Human-Robot Interaction
|
|
PerMIS
Workshops (Performance Metrics for Intelligent Systems)
|
|
