Slide 1
Slide 1 text
Sustainability at Stolt Nielsen
Joost Dooren
September 2023
Slide 2
Slide 2 text
2
Sustainability at Stolt Nielsen
Future generations
September 23
Slide 3
Slide 3 text
3
Sustainability at Stolt Nielsen September 23
Our mission, a bold statement
Emissions target
Emissions target
Carbon Neutral Business
Aspirational target
We aim to have one
carbon neutral ship in
operation.
At least 50%
2030 2050
Carbon Intensity Reduction
– Relative to 2008 levels
Slide 4
Slide 4 text
4
Sustainability at Stolt Nielsen
No easy transition
E.g. Stolt Pride (39000 DWT)
Rough estimation of energy
consumption for Rotterdam –
Houston?
Many variables, e.g. speed
September 23
Slide 5
Slide 5 text
5
Sustainability at Stolt Nielsen
Machine learning models on
environmental and ship metrics
Enables to optimise, anticipate,
innovate and transform
MLOps Platform
September 23
Slide 6
Slide 6 text
6 | Copyright © HSO 2023
Sebastiaan Oude Groeniger
Consultant Data & AI
Slide 7
Slide 7 text
Our goal
A common platform that can be used by all
teams working on different machine learning
use cases
Slide 8
Slide 8 text
Platform requirements
Going to
production faster
Common set of
tools
Common set of
tools
Slide 9
Slide 9 text
Considerations
Several external
suppliers
Align initiatives and
scale knowledge
Different divisions
with their own
experts and
initiatives
Slide 10
Slide 10 text
Key Factors to Success
Centralization to
improve
manageability
Scalable and
adaptable
Keep flexibility for
developers
Slide 11
Slide 11 text
Organise and facilitate the
machine learning process
Organise
Facilitate
Facilitate Data Scientists
Model training
Delivering the right
frameworks, tools
and compute to
focus on model
training
Data
acquisition
Delivering high
quality data
Support
Unblock them to
keep focus on their
job
Slide 12
Slide 12 text
Organise and facilitate the
machine learning process
Organise
Facilitate
03. Model
deployment
04. Monitoring
01. Data
05. Retraining
02. Model training
02
03
04
01
05
Slide 13
Slide 13 text
Implementation
Data Machine Learning API Management Users
Slide 14
Slide 14 text
Azure Machine Learning
Managed endpoints
For model
deployment and
creating API’s
Notebooks &
compute instances
For model training,
managed compute
and safe access to
data
Azure Monitor
To monitor model
and endpoint
performance
Azure DevOps
For automation,
CI/CD and code
repositories
Slide 15
Slide 15 text
15
Sustainability at Stolt Nielsen
Use your data to the fullest!
Let’s give Pieter’s
generation a bright future.
Thank you!
September 23
Slide 16
Slide 16 text
Want to know more?
Come and
visit us at
booth 157