Scribd
Upload
81 views9 pages

Demo

The document discusses using machine learning to determine optimal routes for transporting goods from warehouses to retail stores. Input data like distance, time, weight, and road conditions would be used to continuously learn and identify the best routes that minimize transportation costs and time. As conditions change, the machine learning model would dynamically adjust routes. For example, it could consider all possible routes to deliver items from Warehouse A to Stores C and G, returning to A, and pick the shortest overall route.

Uploaded by

Gaurab Banerji
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
81 views9 pages

Demo

The document discusses using machine learning to determine optimal routes for transporting goods from warehouses to retail stores. Input data like distance, time, weight, and road conditions would be used to continuously learn and identify the best routes that minimize transportation costs and time. As conditions change, the machine learning model would dynamically adjust routes. For example, it could consider all possible routes to deliver items from Warehouse A to Stores C and G, returning to A, and pick the shortest overall route.

Uploaded by

Gaurab Banerji
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 9

SAPATHON 2020

Idea Title- Cognitive Route Determination using Machine Learning for


time and cost of transportation improvement
Submitted By: Gaurab Banerji
Cognitive Route Determination
SAPATHON 2020 IDEA:

Cognitive Route Determination using Machine Learning for time and cost of transportation improvement.

Goods transported from the warehouse to the retail store involves transportation. The time taken and distance
between each retail store/warehouse can be computed to find the best possible route through machine learning. The
dynamic nature of machine learning will also help calculate the optimal routes each time depending on mileage, total
volume of freight, weight, road conditions and traffic. This will help in reduction of fuel costs, faster deliveries from
warehouse to the store. Further, freight size, weight will help calculate the mileage better, which will prioritize which
retail store to visit first for minimizing fuel expenses.
Business Requirement

Business Requirement:
Movement of goods from the Production Plant to the Warehouse, from the Warehouse to the Store, or in between
Warehouses incurs cost of Transportation. This cost needs to be minimized by an automated system which will
determine the best route based on real time data in order to deliver goods quickly while saving transportation costs.
Machine Learning

Input data can be Manual, Automated, GPS Enabled Device, etc.


The program learns from data and recognizes pattern.
It dynamically modifies, tweaks itself to provide the best results.
It is able to take decisions based on observed changes.
More the data, more it learns, and provides better predictions.
The Learning Process

Existing
Actual Data is
(distance,
recordedtime,
afteretc)
travelling
are used
through
to
Actual determine
Data is merged
the with
theoptimalExisting
route. route. Data and
new optimal route is derived.
Route Determination Logic

Similar Logic to Dijkstra’s Algorithm.

New Routes may be added at any time.


Existing routes may be closed due to various reasons.
Roads may become slow due to traffic, weather,
repair work, etc.
Determination of route will change with change in
Route Data.
Simple Example (using Distance only)
The Map on the left shows:
* Location
* Distance
* Direction
A
B J
5 4 4
4 3 3
3 2 2
C D E F

3 6 4
8
H K
G 5
9 2

I
Report Name: ZTT_ROUTEFINDER
Simplify

Suppose a truck at warehouse A wants to deliver items to C and G and


return back to A

The 2 possible routes in this case are A -> C -> G -> A and A ->G -> C -> A.

Here Source and Destination are fixed and the rest of the locations are
arranged using Permutation.

So we need to find the shortest route from A->C,C->G,G->A and A->G,G-


>C,C->A.

Out of these 2 routes whichever is shortest is the Best Possible Route.

Report Name: ZTT_PERMUTATION


Why HANA?

• Each Point is a Location


• 2 Locations are connected
by 1 Unique Path
• Forward and Backward
paths are different as it
could be one way street.
• Even a small distance of 10
KM can have over 1000
different locations and paths
in between depending on
Locations: the details of segregation in
A,B,C,D,E,F,G,H,I,J,K,L,M,N,O (15) location and paths.
Paths
• Huge amount of data needs
AB,BC,CD,CG,DC,DE,EF,EI,FE,FO,GC,
GH,GK,HG,HI,HL,IE,IH,IL,JB,JK,JN,KG to be handled.
,KJ,KL,MA,MN,NM,NJ,NO,ON,OF
(32)

You might also like