KOÇ UNIVERSITY IE FINAL PROJECT -2

By tkilicaslan15 | Updated: May 21, 2020, 8:45 a.m.

Slideshow Video

SHARE THE AWESOMENESS

This video is created for Koç University Industrial Engineering Design Project Presentation. The Team: Binnur Şahin M. Nuri Ulusam Tutku Kılıçaslan Yasemin Savaş

Education

create your own animated videos for free

_abc cc

* Powtoon is not liable for any 3rd party content used. It is the responsibility of each user to comply with 3rd party copyright laws.

Embed code

Select a size, the embed code changes based on your selection.

_abc Powtoon Transcript

INDR 491 
 Industrial Engineering Design Project Presentation
Dishwasher Drying Performance Prediction At ARÇELIK
Faculty ADVISOR :
Assoc. PROF. MEHMET GÖNEN
YASEMIN SAVAS
M.NURI ULUSAM
TUTKU KILIÇASLAN
THE TEAM
Industrial 
Engineering 
Economics
BINNUR sahIN
Industrial Engineering 
Economics
Industrial Engineering 
 busıness admınıstratıon
Industrial Engineering
 socıology
12 BRANDs
17 R&D CenterS
multinational household appliances manufacturer

countries 
+146
digital, smart, and life enhancing
Company Advisor 
Dr. Yusuf Koç
dishwasher drying performance
prediction with 
machine learning tools

Central R+D Center
Laboratory Experiments to Test Drying Performance 
WASHING
RINSING
DRYING
EACH EXPERIMENT TAKES  
3,5 
HOURS
EACH EXPERIMENT IS REPEATED  
5
TIMES
~13000 rows
&

9 COLUMNS
ARE RECORDED
DRY ? -> 2
SEMI DRY ? -> 1
NOT DRY ? -> 0
European Standard EN 50242
+
+
+
=
92 / 100
DR. YUSUF KOÇ
HOW CHANGES
 AFFECT  THE 
PERFORMANCE???
REPEATED EXPERIMENTS 
DIGITIZE THE EXPERIMENTS
SAVE  TIME AND RESOURCES
PROBLEM
IDEA
PROJECT OBJECTIVE
PARAMETERS
PREDICTED
TARGET

MACHINe 
LEARNING
TARGETED 
RESULTs
OPTIMIZED
PARAMETERS
REVERSE
ENGINEERING
METHODOLoGY
EXperiment Results
PRE-PROCESS
TIME SEGMENTATION
MAIN WASH
FIRST RINSING
DRYING
PRE-WASH
SECOND RINSING
13000 rows
x
9 Columns
x 
HUNDREDS ExperimentS
1247 COLUMNS
x 
HUNDREDS ExPERIMENTS
FEATURE
ENGINEERING
REDUCE THE DIMENSION
METHODOLoGY
EXperiment Results
PRE-PROCESS
FEATURE EXTRACTION
TIME SEGMENTATION
METHODOLoGY
PREDICTION
XgBOOST
RANDOM FOREST
SUPERVISED REGRESSION MODELS
CHOOSE 
THE BEST MODEL
PROCESS OPTIMIZATION

KOÇ UNIVERSITY IE FINAL PROJECT -2

Copy this link to share with friends and colleagues:

WELCOME TO POWTOON!