Optimizing Techniques –
Prescriptive Analytics
[email protected]
KP2WFEJ3RX
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action.
Modeling in Prescriptive Analytics Concept
Model
[email protected] Data
KP2WFEJ3RX
Data Analysis
Result
implementation
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action. 2
[email protected]KP2WFEJ3RX
This file is meant for personal use by [email protected] only. 3
Sharing or publishing the contents in part or full is liable for legal action.
Prescriptive Analytics Continuum
[email protected]
KP2WFEJ3RX
This file is meant for personal use by [email protected] only. 4
Sharing or publishing the contents in part or full is liable for legal action.
Prescriptive Analytics in Supply
Chain, an example
• Supply Chain Analytics refers to the improvement in the operational efficiency and
effectiveness by enabling data-driven decisions at
– strategic (≥ 3years to ≤ 5 years),
– operational (≥ 1.5 years to ≤ 3 years),
– and tactical levels (≥ 3 month to ≤ 1.5 years).
[email protected]KP2WFEJ3RX
• It encompasses virtually the complete value chain: sourcing, manufacturing,
distribution and logistics.
• Managing the trade-offs of inventory, customer service and costs of supply chain
source, make and deliver needs to be done but the systems are not aligned to help
managers to understand the inter-connectedness.
• While there are dashboards and reports that show the numbers about what is
happening there is no why—the understanding of the inter-related nature of what
is possible based on the potential of the supply chain.
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action. 5
The Aggregate Planning Model
[email protected]KP2WFEJ3RX
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action.
Requirements
• Forecast Ft for each time period, t in a planning horizon that extends over
time period T
• Production Costs,
– Labour costs – regular time ($/hr) and overtime costs ($/hr)
– Subcontracting cost ($/unit or $/hr)
[email protected] – Cost of varying capacity,
KP2WFEJ3RX
• Cost of hiring or layoff of workers ($/worker)
• Cost of adding or reducing machine capacity ($/machine)
– Labour / machine hours required per unit.
– Inventory holding costs ($/unit/period)
– Stock out or backlog cost ($/unit/period)
• Constraints – Limits on all the above
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action. 7
Outputs of the Model
• Production quantity from regular time, overtime and
subcontracted (time).
• Inventory held.
• Backlog / stock out quantity.
[email protected]KP2WFEJ3RX • Workforce hired / laid off.
• Production Capacity increase / decrease
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action. 8
Costs for Red Tomato Tools
Item of Cost Cost
Material Cost $ 10 per unit
Inventory Holding Cost $ 2 /unit/month
Stock-out / backlog cost $ 5 /unit/month
[email protected] Hiring and training cost $ 300 per worker
KP2WFEJ3RX
Layoff Cost $ 500 per worker
Labour hours required 4 hours per unit
Regular time cost $ 4 per hour
$ 4 x 8 hours/day x 20 days = $ 640 per worker
Overtime cost $ 6 per hour
Cost of Subcontracting $ 30 per unit
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action. 9
[email protected]KP2WFEJ3RX
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action. 10
To conclude the difference between Predictive
and Prescriptive (SCA) analytics
• Difference in the Outcome and thus the
Usage of the analysis.
• Predictive analytics provides you the
Raw Material for making informed
Decisions. For e.g. in the field of
aviation, predictive analytics can use the
[email protected] large amounts of real time performance
KP2WFEJ3RX data to pick up warning signals of
potential component failure.
• Prescriptive analytics provides you with
data backed decision options that you
can weigh against each option. In this
case the options to address the
potential component failures (&
thereby selection of the best option)
are made available.
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action. 11
