3 Reasons why Self-Healing Supply Chains are finally in sight

Self-Healing supply chains are not new.

Exactly 20 years ago, I was leading the deployment of a Self-Healing Supply Chain in a large manufacturing company.

Our BI application tracked the variability of all input and output variables impacting our planning process, predicting with X- and MR- Control-Charts, which potential outliers, would deteriorate the quality of our plans.

Take planning lead times as an example. For every planning lane, we would collect the shipment history, use the actual departure and arrival dates to calculate deviations, and flag the ones requiring a correction.

Automation was only one step away, but twenty years later, we seem to have made little progress.

An intelligent approach to master data management is a key enabler for supply chain excellence, so why is progress in this area so slow?

1. Our lack of focus to apply technology solutions to tackle the issue

Complexity means distracted effort. Simplicity means focused effort. ('Edward de Bono')

For a long time, we preferred to talk about 'Garbage in, Garbage out' to highlight the importance of clean master data, putting the responsibility of maintaining master data on the planners, so we could tackle more exciting problems like multi-echelon inventory optimization or predicting the elasticity of future promotions.

2. Technical and integration complexity stood in the way of a holistic approach

'The biggest challenge is to... have the discipline when there are so many competing things' (Alexa Hirsch)

The data to feed a planning system of records for supply chain planning used to be spread across a realm of applications and technologies and integration concepts, making it hard to develop a single concept to tackle the issue. Recent progress towards centralized supply chain data models makes centralized and efficient master data management much easier than it used to be.

3. Artificial intelligence was not mature enough to provide help but will soon make this a problem of the past

Artificial intelligence is a tool, not a threat. ('Rodney Brooks')

With the emergence of machine learning, and our increasing ability to handle massive datasets, solutions are in the making, self-healing supply chain capabilities will become a reality for everyone.

Estimated Times of arrival, for example, can now be tracked along their journey and the estimated time of arrival adjusted at each step. A delay at customs in the export country and its knock-on effects will automatically adjust the latest Estimate Time of Arrival (ETA) at the destination and add the correct reason code for the delay in the memory of the AI to estimate the ETA date of the next shipment.

In an environment more prone to disruptions, the quality and timeliness of logistic information will have an increasing impact on our plans and business decisions.

No more surprises, no more planning without taking disruption and delays in logistics into consideration.

We needed this new technology to mature to make manual maintenance of product masters, and network nodes but also the supplier sourcing options, routings, and bills of material for new product versions, non-value-adding tasks planners used to do in the very distant past.

Hope is finally in sight!

Artificial Intelligence is what we don't know how to do yet. ('Alan Kay')