9 Mar

How Data And AI Can Revolutionise The Supply Chain

Deepak Vensi

Globally, supply chains are at breaking point. 

Fundamental weaknesses in traditional supply chain management are being brutally exposed by serial crises and novel pressures that are turning up the heat on businesses. 

The Covid-19 pandemic. The war in Ukraine. The cost of living crisis. Rising interest rates. Pressures to reduce carbon emissions. Complex regulatory requirements. 

These pressures are disrupting all aspects of the supply chain, pushing up the cost of raw materials and forcing companies to find new partners because existing partners cannot fulfil the demand. 

As we will see, existing supply chains are too rigid and inflexible. They work acceptably in times of abundance, but start to collapse as soon as demand is constrained or uncertainty sets in. 

As SAP board member Scott Russell commented: “The chaos and the disruption is proving that the supply chain just wasn’t as robust as they expected it to be.

In the face of these pressures, supply chains are falling apart. According to Forbes’ research, the number one priority among global supply chain executives over the next 12 months is reducing supply risk for their materials.

The situation is forcing supply chain executives into a catch-22: in order to ensure fulfilment, they have to hedge their bets, ordering more than they need (“just in case”). This puts more pressure on an already-broken system, which forces even more hedging and breaks the system further. 

The result is waste, inefficiency and even total failure. 

And companies are starting to turn to technology⁠—in particular data and AI⁠—for the solution. The good news is that there are also huge potential business gains to be had from digitising the supply chain and improving connectivity across it: greater transparency, a massive boost to efficiency, huge reductions in waste (both in terms of materials and labour) and, ultimately, a faster and better service for customers

That’s why the big players like AWS and Microsoft are rushing to invest in the next generation of supply chain technology, e.g. AWS Supply Chain and Microsoft’s Supply Chain Platform

In this blog, I will explore how we can use data and artificial intelligence (AI) to make the supply chain more dynamic, resilient and fit-for-purpose in the modern data-driven world.

The Limits of Traditional Supply Chain Management

Supply chain management encompasses the flow of goods and services between businesses and locations. The diagram below shows a typical supply chain, covering everything from the initial raw materials to the long-term customer support with repairs.

Before we look at the solution to the supply chain crisis, we must examine the problems. What are the fundamental flaws in the supply chain that are being so overtly exposed by the recent crises?

There are three main limits:  

  1. Highly complex and manual 

Let’s say your company makes cars. A supply chain for a single vehicle is enormously complex, with many moving parts. From procuring thousands of different raw materials and building the finished product to warehousing the stock and fulfilling customer orders…there are thousands of different parts, partners, places and processes involved.

Yet, most of these parts are still operated manually. Even with pen and paper in some cases! 

Many complex, manual handovers between supply chain
functions that can cause delays and problems

According to Forbes, over 80% of respondents report that they cannot digitally track the movement of direct and indirect materials across their enterprise network.

A lot of time and energy must be invested in passing information manually from one part of the supply chain to another, causing delays and making advanced data work (reporting, analytics etc.) nigh on impossible.

  1. Data is low-quality, siloed and disjointed

According to Forbes’ survey, poor data quality, outdated technology and disparate data silos are the top three causes of inefficiency in the materials management process.

Data quality is low because the emphasis is on capturing the data, rather than using the data. Huge volumes are captured, but they are stored centrally in raw formats that are difficult to use or make available to others in the supply chain.

This is partly because the technology is not fit-for-purpose. Popular (but outdated) supply chain software like SAP is designed to be a system of record, not to provide a data-driven user experience!

Where data is being used, it is still siloed. There is a data module for orders, distribution, fulfilment, tracking and so on, but none of these are integrated! 

Together, these represent a huge barrier to the free flow of (useable!) data along the supply chain. For example, 43% of executives reported a need for greater visibility into inventory availability as the main barrier to sharing critical materials with other parts of their network. 

This creates huge amounts of unnecessary busywork between departments as they collect, send and interpret data (in the form of orders, information and so on) endlessly between each other. 

  1. Producer-centric, not user-centric

The whole system is designed with the producer in mind (i.e. the team at each stage of the supply chain producing the  parts, e.g. by making seat covers and sending that to the team that builds the seats, which would be the ‘customer’). 

In other words, when a particular part of the supply chain is doing their thing, they are only concerned with their work. They aren’t considering what other functions  in the chain might need to know about their operation.

For example, the ESG department might need to report on the carbon footprint of  the production of each car. But is every part of the supply chain making that data available proactively? Or does the ESG department need to spend months badgering hundreds of different people to get the data they need? 

And by the time this data is gathered, the company cannot do anything to optimise their supply chain and reduce their carbon emissions, but only report on it.

There’s limited transparency and communication, which forces people to perform a lot of legwork to gather the data they need to do their jobs. 

3 Ways Supply Chain Management Needs To Change

These three limits mean that your average supply chain is disjointed and disharmonious.

The lack of connectivity, communication and transparency is almost perfectly designed to maximise waste, minimise productivity and create delays at each stage of the process. 

The result is what we’re seeing in the world: collapsing supply chains and panicking businesses. 

In order to overcome these limits, a fundamental transformation in how supply chains are organised and managed is required. Supply Chain Management (SCM) needs to become data-driven, AI-enabled and customer-centric. 

Let’s take a look at each of these in turn.

  1. Data-driven

The only way to tie your entire supply chain together is by putting high-quality, trustworthy, and highly-available data at the heart of your operation. 

Traditional data paradigms treat data like a technical liability. But it must instead be handled like the mission-critical resource that it is. Approaches like data mesh flip traditional data paradigms on their heads, democratising your data and making it available on a domain basis in the form of data products. 

For example, each module in your supply chain could make its data (about the product manufacturing process, carbon emissions etc.) available as a data product, which other parts could then consume as needed. 

You can use these data products to enhance existing services or foster innovation, such as:

  • Real-time reporting, tracking and sharing with third parties and partners
  • Data-driven business decision-making
  • Real-time dashboards and business intelligence 

This makes your supply chain integrated, intuitive and intelligent

You can conduct what-if analyses [what would happen to my semi-conductors if there is a train strike tomorrow?]), do advanced reporting (e.g. track and reduce costs or carbon emissions) and provide value-add services, such as near-real-time order tracking. 

  1. AI-enabled

Once your data has been democratised, you can start applying AI to critical parts of your supply chain to drive novel insights, predictions and business decisions.

Examples include:

  • Predictive modelling
  • Risk analysis and mitigation
  • Automated supply chain optimisation 

You could even combine predictive modelling with third party data in order to anticipate risks to your supply chain based on political or meteorological events (say) in different countries and adjust your procurement accordingly. 

Another application of AI would be by building Digital Twins, which allows enterprises to build virtual representations of their supply chain using real-time data. Any change in the physical system will lead to a change in its digital representation, and such changes can be tracked to the most minute level. In this way, companies can model (say) their warehouses or manufacturing lines, anticipating failures and proactively mitigating them. 

Using AI means that you can go far beyond what a human can do in terms of optimising things like stock planning, logistics, transport and so on. 

  1. Customer-centric

The ideal SCM system would put the customer (these can be internal employees or B2B partners involved in the delivery of the final product or the end customer) front and centre, making not only the core service/product available, but also the relevant manufacturing/fulfilment data. 

In our example with the seat covers above, this would mean that seat covers can not only be easily ordered, but that the data (carbon emissions, production cost, retail price, delivery times etc.) is publicly available for consumption by other parts of the supply chain. Potential delays can be predicted and the order adjusted as necessary. 

By democratising high-quality data across your supply chain you are setting the stage for a customer-centric supply chain, where customers are empowered to use transparent and automated self-service processes to get what they need. 

Where before customers would just throw a work order or query into a black hole and wait an indeterminate amount of time for a response, the experience is closer to ordering something off Amazon: they can make a request and they are guided through a highly-automated process from end-to-end with clear visibility of estimated delivery times, supplier information, frequent notifications and so on. 

In this way, a huge amount of time and energy is saved.

And your people don’t have to hedge their bets due to uncertainty and put more pressure on a broken system. Instead, they are freed up to focus on value-add work, rather than trying to anticipate failures in the supply chain! 

A Vision For The Future of SCM

Consider how the world of booking a taxi has changed. 20 years ago, there were large manual exchanges of data (‘ringing the taxi company’), no transparency (no availability), slow delivery (you could only book in advance) and no feedback loops (no way of complaining about a rude taxi driver!). 

Now with taxi apps, each interaction has been digitised and automated, making each step in the customer experience smaller, faster and more responsive to change. You can see availability in real-time and instantly book a taxi, update your destination or give feedback. 

The same shift needs to occur in the world of enterprise supply chain management, where each event in the supply chain becomes smaller, faster and more automated so that the system as a whole becomes more flexible, dynamic and responsive.

SCM has been evolving since the 60s - the next step is AI/data

By democratising data, eliminating silos, fostering transparency and communication, deploying powerful AI and making your supply chain customer-centric, you start to bridge the gaps between the different parts of your supply chain (employees, partners and customers).

As levels of transparency and automation increase, the whole supply chain becomes more flexible and dynamic. Each event in the supply chain can become smaller and more precise. So, rather than ordering 10,000 semiconductors in case you run out, you can order them as you need them, using data and AI to track progress and anticipate problems. 

This makes it much more resilient in the face of future crises and economic uncertainty. 

At Mesh-AI, we have helped many enterprise organisations to deliver a bold, data-driven, AI-enabled supply chain strategy. We can help you to put data at the heart of your supply chain operations, making it available across your business.

Latest Stories

See More
This website uses cookies to maximize your experience and help us to understand how we can improve it. By clicking 'Accept', you consent to the use of these cookies. If you would like to manage your cookie settings, you can control this in your internet browser. Find out more in our Privacy Policy