6. 數位轉型核心解方(2)-AI Automation (SEA DEEP)
6
core
container
application
container
SEA DEEP
dataset manager
workflow engine
statistical
engine
AIoT deployment
engine
data explorer
engine
multi-tenant
manager
subject manager
event manager
(scheduler)
MQTT manager KAFKA manager
process/task
controller
GPU
resource allocator
form design
engine
machine learning
engine
deep learning
engine
image processing
engine
core containers
application containers
optical
semiconductor EMS education
chemical energy LED
inference
gateway
(IA)
... ...
model W/F
manager
IA W/F
manager
IA supervisor
non-structured data
processing engine
pre-trainer model
manager
members/UAC
manager
engineer/domain
expert
data
scientist
teacher
of AI course
college
students
community
users
AOI影像深度學習應用
AI教育學習
時間序列預測
半導體晶圓瑕疵檢測
AI學習分享KM
7. 數位轉型核心解方(2)-AI Automation (SEA DEEP)
7
Operation
Intelligent
SEA DEEP
AI Automation
Manufacturing
Intelligent
AI
Academy
Marketing
Intelligent
NLP/OCR
model
Customer
Behavior
workflow
modeling
model
training
model
inference
NLP feature
collection
AI tool/
service
model
workshop
EQP
Automation
AIoT
process
automation
model
inference
model
training
marketing
feature feedback
RPA
Intelligent
process
optimization
model
deployment
demand forecasting
Order Fulfillment
Predictive
Maintenance
O2O
Consolidation
FAB
Automation
Customer
Behavior
Data
Collection
User
Behavior
Analysis
real-time SCP and alarm
Supply Chain
Intelligent
APC/Yield
Improvement
Object
Detection
AI School
pricing
intelligent/
commodity
recommendation
`
AI Community
SEA DEEP
Platform
AI use case
8. SEA DEEP人工智慧平台的快速建置步驟
What is SEA DEEP platform?
8
AI model
parameters setup
and training
create your
AI project
datasets
upload
AI model
evaluation
and release
AI model
deployment
(one button)
1 2 3 4 5
13. 數位轉型核心解方(3)-iRPA (智能流程機器人)
13
SUNLITEAI iRPA設計概念與優勢
Workflow Engine
iRPA
open API
iRPA
enterprise
process node
NLP
kernel algorithms
SEA DEEP
adaptor API
iRPA
IA manager
iRPA
Admin UI
iRPA
Modeling UI
iRPA
Aanlytic/Report UI
iRPA
Automation UI
14. 數位轉型核心解方(3)-iRPA (智能流程機器人)
RPA之AI介入時機
14
historical data
(structured & non-
structured)
actors trigger points
Sub-system
boundries
Domain
knowledges
Interfacing
(API)
regulations
Corner cases
For human
intervention
Exception
handling
Data pre-
processing
RPA
Intelligent
Data
understanding
AI model
design/training
AI model
evaluation
AI model
released
SEA DEEP AI Automation