An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

30
AN ONTOLOGY-BASED INTELLIGENT SPEED ADAPTATION SYSTEM FOR AUTONOMOUS CARS Presentation by Lihua Zhao JIST2014 Lihua Zhao, Toyota Technological Institute, Japan Ryutaro Ichise, National Institute of Informatics, Japan Seiichi Mita, Toyota Technological Institute, Japan Yutaka Sasaki, Toyota Technological Institute, Japan

description

Intelligent Speed Adaptation (ISA) is one of the key tech- nologies for Advanced Driver Assistance Systems (ADAS), which aims to reduce car accidents by supporting drivers to comply with the speed limit. Context awareness is indispensable for autonomous cars to perceive driving environment, where the information should be represented in a machine-understandable format. Ontologies can represent knowledge in a format that machines can understand and perform human-like reason- ing. In this paper, we present an ontology-based ISA system that can detect overspeed situations by accessing to the ontology-based Knowl- edge Base (KB). We conducted experiments on a car simulator as well as on real-world data collected with an intelligent car. Sensor data are converted into RDF stream data and we construct SPARQL queries and a C-SPARQL query to access to the Knowledge Base. Experimental re- sults show that the ISA system can promptly detect overspeed situations by accessing to the ontology-based Knowledge Base.

Transcript of An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Page 1: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

AN ONTOLOGY-BASED

INTELLIGENT SPEED

ADAPTATION SYSTEM FOR

AUTONOMOUS CARS

Presentation by Lihua ZhaoJIST2014

Lihua Zhao, Toyota Technological Institute, Japan

Ryutaro Ichise, National Institute of Informatics, Japan

Seiichi Mita, Toyota Technological Institute, Japan

Yutaka Sasaki, Toyota Technological Institute, Japan

Page 2: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Outline

Motivation

Related Work

Approach

Experiment

Conclusion & Future Work

2

Page 3: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Avoid overspeed to reduce car accidents.

Advanced Driver Assistance Systems (ADAS)

Intelligent Speed Adaptation (ISA): one of the most cost-efficient

way to improve roadway safety.

Enable autonomous cars to perceive driving

environment.

Ontology-based Knowledge Base

Advanced Digital Map: road information, speed limits, etc.

Traffic Regulations

Motivation3

Page 4: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Use ontology and 14 SWRL rules to enable the vehicle to

understand the context information when it approaches road

intersections. [Armand, 2014]

Automation level ontology and situation assessment ontology

are designed for co-driving. [Pollard, 2013]

A complex intersection ontology (car, crossing, road connection,

and sign at crossing) is introduced for fast reasoning. [Hulsen, 2011]

An ontology-based traffic model that can represent typical traffic

scenarios such as intersections, multi-lane roads, opposing traffic,

and bi-directional lanes is introduced. [Regele,2008]

Related Work4

Page 5: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

System Flowchart

Knowledge Base Ontologies

Instances

Rules

Query SPARQL Query

C-SPARQL Query

Ontology-Based Intelligent Speed Adaptation

System5

Page 6: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Input

Sensor Data

PreScan driving simulator

GPS-IMU sensor

Knowledge Base

Ontology-based data

Output

Overspeed warning

System Flowchart6

Page 7: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Ontology: Machine-understandable knowledge representation

Classes: called as Concepts, defined by owl:Class.

Properties: owl:ObjectProperty and owl:DatatypeProperty.

Instances: individuals of a domain, defined by owl:Thing.

Rules: describe logical inferences, with if-then sentence.

Ontology Editor

Protégé ontology editor

Ontologies7

Page 8: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Enable autonomous cars

to perceive driving

environment

to make safe driving decisions.

Knowledge Base

Components

Ontologies

Instances

Rules

Knowledge Base8

Page 9: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Describe road, intersection, lane, and speed limit. (78)

ObjectProperty (18)

map:isLaneOf

map:isRoadSegmentOf

DatatypeProperty (31)

map:speedMax

map:boundPOS

map:osm_ref

Map Ontology9

Page 10: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Describe the path of autonomous cars. (34)

ObjectProperty (15)

control:nextPathSegment: intersection or lane

DataProperty (2)

control:pathSegmentID

Control Ontology10

Page 11: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Concepts of vehicles and devices such as sensors.

(33)

ObjectProperty (3)

car:usedSensor

DataProperty (15)

car:car_length

car:car_ID

Car Ontology11

Page 12: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Instances are also known as individuals that

model abstract or concrete objects based on the

ontologies.

Tempaku Map Instance

Path Instance

Car Instance

Instances12

Page 13: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Map instances include

roads, road segments,

intersections, lanes,

schools, etc.

speed limits

enter & exit of lanes

connection of road

segments

Tempaku Map Instance13

Page 14: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Constructed based on the Tempaku map and control

ontology.

next path segment

Path Instance14

Page 15: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Describe a car and devices installed on the car.

Car Instance15

Page 16: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Semantic Web Rule Language (SWRL) is used to

express rules.

Pellet reasoner is used for ontology reasoning.

E.g.: If a car is running on a road near a kindergarten. The speed

limit should be 30km/h near the kindergarten, even though the

default speed limit is 40km/h on the road.<tempaku:Takasaka_Kindergarten, map:nearTo,

tempaku:Hisakata2RS2>

Rules:

Rules16

Page 17: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

SPARQL Query

A powerful RDF query language.

Access to the ontology-based Knowledge Base.

C-SPARQL Query

Access to the RDF stream data.

Format: <subject, property, object, timestamp>

Queries for ISA system17

Page 18: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Retrieve the next path segment based on

current path segment. (pathSegmentID: 0, 1,

2, …, n)

SPARQL Query I18

Page 19: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Retrieve the speed limit of current path

segment.

SPARQL Query II19

Page 20: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

If a car’s average velocity in the past 500ms exceeds its

own speed limit. (i.e. maxSpeed:120km/h)

RANGE: duration to receive sensor stream data

STEP: frequency of a sensor receiver.

C-SPARQL Query20

Page 21: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Experiment Settings

Knowledge Base

PreScan Simulator Experiment

Real-World Data Experiment

Experiment21

Page 22: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Computer Specification

PreScan driving simulator car and a smart

vehicle

Experiment Settings I22

Page 23: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Trajectory for the experiment (near TTI

campus)

Experiment Settings II

PreScan Map (OpenStreetMap) Google Map

23

Page 24: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Knowledge Base for

Experiments24

Page 25: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Speed: 8 ~ 18 m/s

Smooth acceleration,

deceleration, constant

speed.

SPARQL: 11ms (2 ~

23ms)

Reasoner: 242ms

PreScan Simulator Experiment

Kindergarten

25

Page 26: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Drive the smart vehicle

GPS-IMU sensor

SPARQL: 11ms (3 ~

23ms)

Reasoner: 177ms

Real-World Data Experiment

Kindergarten

26

Page 27: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Advantages Evaluate the ISA system with the PreScan driving

simulator and a smart vehicle.

Retrieve knowledge from Knowledge Base at real-time.

Effectively detect overspeed situations.

Problems Shifts of GPS positions on PreScan driving simulator.

Delays of data transmission with GPS-IMU sensor.

Discussion27

Page 28: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Intelligent Speed Adaptation System

Ontology-Based Knowledge Base

SWRL rules for ontology reasoning

SPARQL and C-SPARQL for knowledge

retrieval

ISA system evaluation with PreScan driving

simulator and real-world GPS-IMU sensor

data.

Conclusion28

Page 29: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Add more knowledge

Traffic light, traffic regulations.

Improve driving safety.

Add links to external resources

Discover hidden knowledge from interlinked

instances.

Future Work29

Page 30: An Ontology-Based Intelligent Speed Adaptation System for Autonomous Cars

Lihua Zhao: [email protected]

Thank you !