8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
1/25
CIS 2011 rkala.99k.org1st September, 2011
Planning of MultipleAutonomous Vehicles usingRRTRahul Kala, Kevin Warwick
Publication of paper: R. Kala, K. Warwick (2011) Planning of Multiple Autonomous Vehicles using
RRT, Proceedings of the 10th IEEE International Conference on Cybernetic Intelligent Systems , Docklands, London, pp.20-25.
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
2/25
CIS-2011, 1st September, 2011 2
Autonomous Vehicles
Safety
EfficientDriving
JamAvoidance
Coordination
Comfort
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
3/25
CIS-2011, 1st September, 2011
Software Architecture
SensorEnvironment
understanding
Sensor
fusionLocalizationPlanning
Control Motion
Map Mission
Environment
3
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
4/25
CIS-2011, 1st September, 2011
Software Architecture
4
Coordination
Vehicle 1
Vehicle 2
Vehicle 3
.
Vehicle n
Inter-Vehicle
Communication
CoordinationObstacle Discovery
Localization
Collision Avoidance
Travel Plan
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
5/25
CIS-2011, 1st September, 2011
Junior (Stanford) Model
5
Planning
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
6/25
CIS-2011, 1st September, 2011
Why not speed lanes?
6
Highly diversesizes
Excess road
widths with
diverse speeds
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
7/25CIS-2011, 1st September, 2011
Why not speed lanes?
7
Single lanes
Complex obstacle
framework
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
8/25CIS-2011, 1st September, 2011
Talos (MIT) Model
8
Key Takes:
Planner/Controller
integrated
Possibility of multiplevehicles
Source:Y.
Kuwata,S.
Karaman,J
.Teo,
E.
Frazzoli,
J.P.
How,
G.
Fiore,
Real-TimeMotio
nPlanning
With
ApplicationstoAutonomousUrbanDriving,
IEEETran
s.ControlSyst.Technol.,vo
l.17,no.5,
pp.11
05-1118,
2009.
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
9/25CIS-2011, 1st September, 2011
PLANNING ALGORITHM
9
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
10/25CIS-2011, 1st September, 2011
Layers of Planning
10
Route Planning
RoadSegmentation
Road SegmentPlanningAbstraction
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
11/25CIS-2011, 1st September, 2011
Overlapping Segment Planning
11
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
12/25CIS-2011, 1st September, 2011
Coordinate Axis System
12
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
13/25CIS-2011, 1st September, 2011
RRT
13
Source and
direction
Goal
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
14/25CIS-2011, 1st September, 2011
Curve Generation
14
max,min vkv
)(2)()( ddtdik
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
15/25CIS-2011, 1st September, 2011
Coordination
Principle: Prioritization
Rule: Higher priority for vehicle earlier in
segment Communication: Trajectory followed
stored as hash maps
Speed reduced in steps of till feasibleplan found
15
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
16/25CIS-2011, 1st September, 2011
Results
16
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
17/25CIS-2011, 1st September, 2011
RRT Generated
17
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
18/25
CIS-2011, 1st September, 2011
Vehicle Following
18
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
19/25
CIS-2011, 1st September, 2011
Overtaking
19
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
20/25
CIS-2011, 1st September, 2011
Vehicle Avoidance
20
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
21/25
CIS-2011, 1st September, 2011
Analysis
21
0
100
200
300
400
500
600
700
800
10 20 30 40 50 60 70 80 90 100 110 120 130
Numberofnodesintree
Step Size
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
22/25
CIS-2011, 1st September, 2011
Analysis
22
0
500
1000
1500
2000
2500
10 20 30 40 50 60 70 80 90 100 110 120 130
Totalnumberof
iterations
Step Size
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
23/25
CIS-2011, 1st September, 2011
Why RRT?
* Assuming high resolution maps, traffic scenario, complex obstacle framework.
Reduction of resolution not possible
23
Property Algorithm Order*
Completeness Graph > RRT > Evolutionary > Behavioral
Ability to generate paths
with large number ofobstacles
Graph > Behavioral > RRT > Evolutionary
Path length/ Optimality Evolutionary > Behavioral > RRT > Graph
Computational Cost Behavioral < RRT < Evolutionary < Graph
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
24/25
CIS-2011, 1st September, 2011
Conclusions
24
Problem
Planningmultiplevehicles
Different timesof emergence
Traffic Scenario
Constraints
ComputationalTime
ComplexObstacleFramework
Solution
RRT Road Axis
System
CurveSmoothening
Priority
Open Issues
Optimality Being un-
struck
Cooperation Non-
autonomousvehicles
Uncertainties Curves
RRT-Connect
8/4/2019 Planning of Multiple Autonomous Vehicles using RRT
25/25
CIS 2011 1st S t b 2011
THANK YOU
Acknowledgements:
Commonwealth Scholarship Commission
in the United Kingdom
British Council
25