Optimal routing development based on real voyage data presented by_sewonkim

2017.Nov.2
Korea Japan Smart ship Joint Session
Optimal Routing Development
Based on Real Voyage Data
DSME/KMA/HMM
S.W.KIM, J.W.CHOI, H.R.PARK, D.J.JUNG, S.S.Byeon, H.M.EOM
SNAK Smart-Ship Joint Session
NOV.2017.2ND
1

Takeaway
▪ The state of art:AutonomousVessel
▪ A Key Challenges : Routing Optimization
▪ Our Approach
▪ Realistic Environment Modeling
▪ Precise Performance Estimation
▪ Power Comparison based on RealVoyage Data
▪ Case Study: East Bound Container
2

Recent News
• Amazon,Alibaba-Mearsk
• Rolls-Royce and Google / ABB and IBM
• Kongsberg andYARA International
• Autonomous Ship on IMO 2017 Agenda
3

2017.Nov.2
Korea Japan Smart ship Joint Session
Future is already here
- Global movements for developing autonomous ship-
ABS ROYAL NAVY
MUNIN,AAWAI,YARA:2019
MUNIN : Maritime Unmanned Navigation Intelligence in Networks
AAWAI : Advanced AutonomousWater-bone Application Initiative
JAPAN : NYK plant to launch Autonomous Container Carrier to 2019
CHINA : Shipyard – Shipping Industry – Research Institute Alliance
US : Remote OSV operation completed in 2017
Korea
NYK 201
CHINESE 2020NAVY Korea
5

Why Autonomous?
25% of CAPEX
20% of OPEX
Training Cost
96 % Ship Collision1)
PiracyVictim
1) Dr. Rothblum, Human Error and Marine Safety, USCG, 2012
6

Challenges Towards
AutonomousVessel
Trim
Optimization
Fouling Smart
Monitoring
Remote
Control
Cost -Efficient
Voyage
IAS
Platform
Port
Automation Logistics
Optimization
Risk
Analysis7

Cost EfficientVoyage
- Technology Roadmap-
Bad Weather Avoidance
FOC(Fuel Oil Consumption)
Response Based/Multi Objects
Route Optimization
8

Information Optimization Decision
Optimization Frame
(Input)
Weather, Ship..
(Evaluation) (Output)
Optimal
Route
Performance
9

Performance Estimation
Resistance, Motion
Weather,Trim, Fouling
Propeller Emerging
Engine Dynamics
10

Object (ex: FOC)
Constraints
Piracy
Cargo Safety
Land Avoidance
Arrival Time
Ice
Fuel Factors
Calm Resis.
Wave Added
Wind Load
Generic Input
Sea Chart
Weather Forecast
Case Input
Target
Fuel Consumption
Ship Response
Main Parameters
Routing
Optimization
11

Weather
R&D Collaboration
Ship Voyage
“Goal : Realistic
Performance Estimation”
12

• Resistance
• Wind
• Wave
Increasing
Power
Consumption
13
Major Factors of
Power Consumption

Increased Power
     _
_
, ,
, ,
: , ,
:
:
calm wave added wind
Power R U R U R U UT
R R R
Calm Water Wave Added andWind Resistance
U Vessel Speed
Vessel Heading Angle
T VoyageTime
 

    

14

Estimated
Power
(Analysis
+ Model Test)
Measured
Power
(Voyage
Data)
VS
15
Power Comparison based
on RealVoyage Data

Case Study : HMM Hope
LPP: 349.5m
Breadth:48.4m
Draft:14m
Volume : 162517 m3
16

Target : fromYokohama to Sanfrancisco
Period: June 1ST 2016TO JUNE 10TH 2016
Great Circle Distance : 8,206 km
Yokohama San Francisco
17

Routing Optimization Problem
     _
_
,
, ,
:
, , : , ,
:
:
Object to minimize FOC
where
FOC R U R U R U UT SFOC
FOC Fuel Oil Consumption
R R R Calm Water Wave Added andWind Resistance
U Vessel Speed
Vessel Heading Angle
T VoyageTime
SF
 

     

: ( / )
: , , ,
OC Specific Fuel Oil Consumption ton kwh
Suject to VoyageTime Ta rget Position Speed Range Heading Angle Range
18

Optimization Algorithm: Iterative DP
A method for solving complex optimization problem by breaking it down
into a collection of simpler sub-problems using the iterative calculation
based the back propagation optimum theory.
Concept Diagram 19

Routing Optimization Procedure
1) Divide WholeVoyage Routes into Unit Step
2) Load Ship/Weather/Voyage Data
3) Create Speed and Heading Command Seeds
4) Find Optimum which has Minimal Fuel Oil Consumption
• Calculate FOC based on Estimated Power
(Total Resistance)
20

Calm Water Resistance
Calm Water Resistance due to Speed
21 22 23 24 [Knot]
:calmR Calm Water Resistance
21
     _ , ,wave addca inl e w dm dPower R U R UR UTU      

Wind Load Estimation
   2 2
0.5 0.5 0wind A AA WR T WR A AA L GR C A V C A V   
ρA : Air Density
VWR /ψWR : Relative Wind Speed and Direction
AT / VG : The Projected Area and Advancing Speed of the Ship
CAA : Wind Load Coefficients
Wind Load Coefficients : CAA 22
     _ ,,calm wave adde windd RPower R U UR U UT    

Wave Added Resistance
viaWish SNU
Mean Drift Force Estimation based on Potential Theory due to various
Speeds and Drafts
Strip Model RANKINE Panel Model
23
     _ ,,wave addcal inem w ddPower R U R U UR TU      

 
 _ 2
0
,
,wave added
A
QTF
R E d
 
  


 
ω, α, and ζA : Frequency, Direction, and Amplitude of the Wave
E(ω,α): The wave spectrum.
ModelTest under Irregular Waves
24

:Design Draft
Mean Drift Force due to Wave Heading Angle
25

Realistic Weather Forecast Data
by KMA
3 Hour Based
Ensemble Forecast Data
26

Time
Position data
Latitude (radians & degrees)
Longitude (radians & degrees)
Environmental data
Wave height (m), wave period (seconds)
Wave direction (degrees)
Wind speed (knots)
Wind direction (degrees)
Ship data
Speed over ground (knots)
Longitudinal water speed (knots)
Draft Aft, Fwd (m)
Trim Dynamic (m)
Propulsion RPM (rpm)
Propulsion Power (kW)
Propulsion Torque (Nm)
Rudder Angle (degrees)
RealVoyage Data from HMM
27

Power Comparison
Power History
28

Routing Optimization Result
Calculated Route
Calculated Route
29

Conclusion
▪ The optimal FOC routing was conducted by DSME, HMM, and KMA.
▪ The model test results (calm water, wave added, and wind resistance) were
considered to estimate the power increase.
▪ The measured power and the estimated power were compared based on
real voyage data.
▪ The 4% FOC reduction was achieved compared to great circle on eastbound
route for 13.1K container carrier on June 2016.
▪ Lesson Learned :
:A precise performance estimation is a crucial factor for the optimal routing
▪ Next Goal
: Hydrodynamic performance analysis integration
30

AutonomousVessels is not
far away but is closer than
you might think.
31

Thank You So Much!
SEWON KIM
sewonkim@dsme.co.kr
+82-2-2129-3707
32

Optimal routing development based on real voyage data presented by_sewonkim

Recommended

More Related Content

What's hot (20)

Similar to Optimal routing development based on real voyage data presented by_sewonkim (20)

Recently uploaded (20)

Optimal routing development based on real voyage data presented by_sewonkim