The future of... Supply Chain Management

Dear friends,

Working on real-time integrated supply chain is challenging as it touches every corner of the organization and secondly sciences and operational practices clash. Soon you can't see the forest for the trees right from the start.

Stay with me, we can deep dive on some of the typical hurdles on your road to real-time Supply Chain.

Imagine you are a project manager or part of a sub-project team.

Let's dive into REALTIME

For operations, real-time can have a different meaning as to a data scientist or a data engineer though basically we want to have the same outcome. Operations is used to talk about "incidents" the physical action of stalling or moving goods. The information is based on TRUST between parties.

From an IT point of view, real-time can mean based on the transaction time between the sender and the receiver of data which is often measured in milliseconds.

You may think about a clever way how to deal with a acceptable definition before a successful kick-off of your project.

Speaking about technology, SCM becomes more and more support from mathematical models, there also needs to be clarity about Strategy Building and Demand Prediction since we are in a sub-project team. Our SCM heart beats faster when we have to share our knowledge about Inventory, seasonality, warehousing and handling costs, margins. We know how we predict these, following last year's and past 5 years curves and project them into the future, add some slack and deal with some risks. Ready!

No, in the time of digitization we need numbers.

Demand prediction and business strategy processes share several commonalities, particularly in their roles within organizational planning and decision-making.

We want to remain professional, and come up with following models:

Business Strategy Models

• Porter’s Five Forces Analysis:
• Balanced Scorecard (BSC):
• SWOT Analysis:
• PESTEL Analysis:

Evaluates the external macro-environmental factors affecting an organization: Political, Economic, Social, Technological, Environmental, and Legal.

• Scenario Planning:

A strategic planning method that allows businesses to develop and test various future scenarios, preparing them for uncertainties and potential changes in the market environment.

Excellent!

Now have a look at some common demand predict models, forget about that crystal ball they used to discuss in the catacombs of your company. Here are some models to start with predicting inventory /sales maybe transports?

Demand Prediction Models

• Machine Learning Algorithms:

Linear Regression: Predicts demand based on the relationship between dependent and independent variables.

Random Forest: Uses multiple decision trees to predict demand, reducing the risk of overfitting.

Neural Networks: Complex models that can capture non-linear relationships and interactions between variables.

ARIMA (Auto-Regressive Integrated Moving Average): A time-series forecasting method that uses past values and lagged variables to predict future demand.

• Deep Learning Models:

Recurrent Neural Networks (RNNs): Particularly useful for time-series data, as they can learn dependencies and trends over time.

Long Short-Term Memory (LSTM): A type of RNN that can learn long-term dependencies, making it effective for demand forecasting over extended periods.

Convolutional Neural Networks (CNNs): Typically used for image data but can be applied to time-series forecasting by transforming the data into images.

• Hybrid Models:

Combining traditional statistical methods with machine learning approaches can improve forecast accuracy. For example, using ARIMA to capture linear trends and machine learning models to capture non-linear patterns.

SCM has to be spelled and handled differently from now on!

How good is the data you sit on?

Is your data End to End?

Is your data comprehensive enough to incorporate insights about all players, market, risks etc?

Wish you a great day, many learnings, smile.

Abraham