1. AI, ML and Advanced Analytics
Manufacturing – Industrial IioT
Rapidly providing the right data … to your best people …
to enable them to solve your biggest business challenges.
2. ERP CRM Stock eCommerce
HMI
Drive to converge IT & OT data
3. HMI
ERP CRM Stock eCommerce
Historian Data Lake
Seeking ‘Single pane of glass’ - UNS
4. Consolidate &
Contextualise Data
Ongoing Strategic Direction Continually Adding Systems & Equipment
Standard Reporting,
Ad-Hoc Reporting
and Basic Analytics
Raw Data
This only provides an underlying foundation
5. Raw Data
Consolidate &
Contextualise Data
Standard Reporting,
Ad-Hoc Reporting
and Basic Analytics
V
A
L
U
E
Advanced Analytics,
ML – Predictive Analytics
Prescriptive Analytics
Like this, the data cannot be used for AI - ML
6. Raw Data
Consolidate &
Contextualise Data
Standard Reporting,
Ad-Hoc Reporting
and Basic Analytics
V
A
L
U
E
Advanced Analytics,
ML – Predictive Analytics
Prescriptive Analytics
Load ML-Ready Data
& Build Model
Prepare Data for
Machine Learning
Deploy Model
Ongoing Mgmnt
The data must be further ‘Prepared’
7. Raw Data
Consolidate &
Contextualise Data
Load ML-Ready Data
& Build Model
Prepare Data for
Machine Learning
Deploy Model
Ongoing Management
Typically 1 - 2
days effort
Typically 3 - 4
months effort
Typically 2 - 3
weeks effort
This data preparation process is manual and complex
8. Consolidate &
Contextualise Data
Load ML-Ready Data
& Build Model
Prepare Data for
Machine Learning
Deploy Model
Ongoing Management
Typically 3 - 4
months effort
Typically 1 - 2
days effort
Typically 2 - 3
weeks effort
Duration 4 – 6 Months
Data Scientist &
Data Engineer
Salary Ranges
Internal Data Team “85%’ of ML initiatives fail”
Gartner
Raw Data
Project Failure – Complexity, Duration & Cost
Chasm
Min € 60 – 100 K per annum
11. Microsoft
8 Mths
6 Mths
6 Mths
7 Mths
9 Mths
36 Mths – 3 Yrs !!!
Microsoft - Is this really expected IioT project duration?
12. Consolidate &
Contextualise Data
Raw Data
Ongoing Strategic Direction Continually Adding Systems & Equipment
Continue ‘Single Pane of glass’ – UNS strategy
Duration 4 – 6 Months
Data Scientist &
Data Engineer
Salary Ranges
Min € 60 – 100 K per annum
Internal Data Team
13. Consolidate &
Contextualise Data
Load ML-Ready Data
& Build Model
Prepare Data for
Machine Learning
Deploy Model
Ongoing Management
Project durations, typically 5 – 10 Days
Xpanse delivery - Typically 25 - 30% of current project costs
Raw Data
Each AI, ML & Advanced Analytics solution can be rapidly and cost effectively delivered, much shorter time to value
While leveraging AI, ML & Advanced Analytics
Duration 4 – 6 Months
Data Scientist &
Data Engineer
Salary Ranges
Min € 60 – 100 K per annum
Internal Data Team
14. An example of what it can look like when you get to cross the Chasm and leverage powerful Insights
Xpanse – Disruptive solutions crossing the Chasm
16. HMI
Historian Data Lake
600+ Data Tags
ML used to determine which Tags – Data
correlated to the failure event
SME and Engineering team explored
how they could impact the failure KPI
Semiconductor Sector
Failure Process Optimisation
17. We have a failure rate of 20.1% in our manufacturing process, this is what an overview of the data looks like
Xpanse – Goes way beyond basic analytics
18. If we can operate within the parameters highlighted above, we should be able to reduce this failure rate to 9.2%
Advanced Analytics – Data Exploration & Insights
19. Together we workshop the initial scope of needs
We collaborate with the teams that own the problem
on clarifying the requirements.
We deliver, deploy and monitor models/dashboards
If needed - we feed the outputs to your systems
We deploy the Xpanse AI platform and connect to your data
Xpanse – Typical Engagement Model
20. If we can operate within the parameters highlighted above, we should be able to reduce this failure rate to 9.2%
Advanced Analytics – Data Exploration & Insights
Q & A