Center line content
ROBUST LOCALISATION OF AUTOMATED GUIDED VEHICLES IN DYNAMIC LOGISTICS ENVIRONMENTS
Markus Boehning
SICK AG, Central Unit Research
Center line content
IT‘S ALL ABOUT LOCALIZATION AND SAFETY MOTIVATION FOR AUTOMATED GUIDED VEHICLES
Center line content
DEL MONTE PLANT EXAMPLE OF AUTOMATED WAREHOUSE LOGISTICS
Center line content
LOCALISATION APPROACHES FOR AGVS REQUIREMENTS AND SOLUTIONS
Track-/Grid-based navigation • Applications with simple or predefined paths
Reflector-based localisation • Applications with free and ad hoc planned paths
• Uses navigation sensors such as SICK NAV350
• High installation efforts and costs for artificial landmarks
Contour-based localisation • Applications in environments without artificial landmarks
• Either dedicated navigation laser scanner (AGV) or
reuse of safety laser scanner data (AGC)
Center line content
SMART MICRO FACTORY FOR E-CARS WITH LEAN PRODUCTION PLANNING SMART FACE
Center line content
CONTOUR-BASED LOCALISATION IN SMART FACE DEMONSTRATORS
Contour-based localisation using Monte-Carlo localisation (MCL)
SICK S300 Shopfloor-Demonstrator
SICK TiM5xx Minidemonstrator
Center line content
CONTOUR-BASED LOCALISATION APPROACH AND CHALLENGES
• Mobile objects are not registered in the digital environment map
• Low accuracy or failure of the MCL
Center line content
LOCALISATION IN DYNAMIC ENVIRONMENTS UPDATE OF THE REFERENCE MAP
• Contour-based localisation using MCL
approach
• Evaluation of pose accuracy based on
Measurement/outlier ratio
Variance of position estimations
• Map update after repeated observation of
position with high accuracy
Localisation
Accuracy evaluation
Map update
yes no
Center line content
LOCALISATION IN DYNAMIC ENVIRONMENTS UPDATE OF THE REFERENCE MAP
Center line content
LOCALISATION IN DYNAMIC ENVIRONMENTS UPDATE OF THE REFERENCE MAP
• Comparison of position and angle accuracy
with dynamic map update (blue)
without dynamic map update (red)
• Ground truth reference
reflector-based laser localisation (NAV350)
Center line content
Markus Boehning
SICK AG, Central Unit Research
MANY THANKS FOR YOUR ATTENTION.