Getting to grips with AI-driven supply chains
CAN you imagine a world where artificial intelligence (AI) can transform logistics to such an extent that the supply chain establishes new standards for productivity and efficiency?
Accenture Digital ASEAN’s Consulting Director of AI, Rohit Dhawan can. According to him, an AI-driven supply chain is one that fully exploits the potential of new AI and machine learning technologies. For each touchpoint in a supply chain, AI is used to either optimize current operations or simulate future scenarios.
In an exclusive interview with Tech Wire Asia, Dhawan explained what AI-driven supply chains really do and how they benefit businesses.
If you think of the decision points in a supply chain and the interdependencies between them, algorithms that learn and simulate from current data are driving several use cases. Be it forecasting, planning, simulation, network optimization, or intelligent scheduling, it can all be driven by AI.
“In an ideal scenario, a machine should be able to run your supply chain making the best use of available resources and maximizing your throughput,” explained Dhawan.
However, the reality is that supply chains differ across industries so applications will depend on the nature of the supply chain and the business context.
For instance, in telecommunications, supply chains may be used to install new broadband connections, whereas in mining, they may be used to produce and move finished products from raw iron ore.
Regardless of the industry, supply chains work on fundamentally core principles of movement of goods or people across a series of nodes.
“Given the amount of data that is captured in this movement and the availability of new digital tools and techniques, AI has evolved as a potent topic when it comes to injecting the “smarts” across the chain,” claimed Dhawan.
Businesses see the benefit
An increasing number of organizations are adopting digital technologies in their supply chain.
AI is already helping these organizations gain a better understanding of demand and supply variables and helping them anticipate and plan “what if” scenarios.
The ability to simulate provides planners and decision makers the data and facts to make (or not) specific investments and allocations. It also reduces uncertainty and risk.
The other benefit of an AI-driven supply chain is increasing the effectiveness of decisions and their subsequent automation.
“As data accumulates, AI algorithms are able to “learn” from history and provide decision recommendations that can take the form of a forecast, a scenario or a recommendation,” suggested Dhawan.
The business benefits are in the form of decreased supply chain costs and increased process efficiencies.
Dhawan cited the example of retailers using AI or machine learning to analyze demand variations and trends in the fashion market to gain a more accurate view of matching supply to demand.
Autonomous fork-lifts in warehouses and driver-less autonomous vehicles for deliveries are also a near reality, he pointed out.
“Overall, the adoption of AI in supply chain planning and operations is still in its infancy. Lack of understanding of this domain coupled with a lack of skills to implement has deterred organizations to rise on the maturity curve,” explained Dhawan.
There are examples of companies using AI for specific supply chain problems such as inventory optimization or network planning.
However, more mature organizations need to exploit AI across the full chain, taking into consideration the interdependencies between the links thus getting a “whole of system” view.
How to get started
The first step to implement AI within a supply chain is to ensure that AI is a key component of the broader digital strategy and that optimization of the supply chain has been recognized as a key value driver for the organization.
One of the key drivers of adoption is executive sponsorship – it all starts at the top with the right strategy and vision.
According to Dhawan, decision-makers must understand what AI is and how it can generate benefit for them to invest in this technology.
Since the implementation of AI is still quite contextual and bespoke, implementation is not straightforward and it will be a while before the use of AI becomes commoditized and commonplace.
“To start on the journey, the supply chain needs to broken-down into its nodes and touchpoints to gain a holistic understanding of how value is generated,” said Dhawan, who is speaking more on the topic at the ConnecTechAsia Summit in Singapore this week.
However, there are still a lot of organizations that do not have an accurate picture of their supply chain processes.
For each touchpoint, an AI-expert can establish the use cases that can provide value for either planning or operations.
Once the use cases have been identified, they can be prioritized according to business benefit and ease of implementation.
Execution of an AI project must coincide with other digital initiatives as AI will typically leverage the same underlying data infrastructures, process and subject matter expertise.