How can AI be used in supply chain management?

It is a simple fact that supply chain waste costs money and needlessly occupies valuable resources. Global issues such as Covid-19, the war in Ukraine, Brexit and related border blockages, have caused international mayhem to supply routes. But localised, day-to-day disruption also causes delays, risking profit and cashflow.

Over the last couple of years, Artificial Intelligence (AI) has come of age and is now penetrating every business sector. It will have (and already is having) a profound impact on the logistics and supply sector. It is helping businesses streamline logistics operations and improve efficiency in their supply chain management. However, as the technology becomes increasingly integral to these services, companies must also be aware of the challenges and risks of AI.

Benefits of using AI in logistics

AI’s fundamental strength is the ability to use learned logic, the process of applying knowledge from past training and experience, to solve complex logistical problems. Traditional systems lack this ability as they rely on established rules and practices without the emphasis on learning and evolving through experience. AI is therefore better equipped to analyse unstructured data, recognise patterns and make informed decisions.

A 2022 report by Avery Dennison found that globally about 8% of stock perishes or is discarded annually, costing$163bn. While this is bad for business it also has a huge impact on the environment.

A 2022 report by Avery Dennison found that globally about 8% of stock perishes or is discarded annually, costing

$163bn

. While this is bad for business it also has a huge impact on the environment.

Replenishment Optimisation

Replenishment optimisation-focussed AI algorithms, such as Continuous Replenishment Program (CRP) and Economic Order Quantity (EOQ), can leverage past and present data to anticipate future stock replenishment needs far more quickly than a human can, helping to reduce wastage and reduce “black orders[1]”.

Replenishment optimisation-focussed AI algorithms, such as Continuous Replenishment Program (CRP) and Economic Order Quantity (EOQ), can leverage past and present data to anticipate future stock replenishment needs far more quickly than a human can, helping to reduce wastage and reduce “black orders

[1]

”.

Routing Optimisation

AI can also help solve routing management problems.

The shift toward home working during the pandemic has altered distribution patterns, and companies have had to adapt to an increase in online orders as people stay at home. At the same time, companies have had to cope with a reduced workforce as their vulnerable and sick employees stay home too. AI can use past and present data to improve delivery punctuality and predict which deliveries are likely to be late or early.

The algorithm can be tailored to specific company delivery locations, vehicle capacity and can take into account traffic issues. It can also consider individual constraints, such as timeframes, and company objectives, including maximising the number of deliveries per route or optimising the number of delivery stops to ensure businesses are using their resources and employee time efficiently.

Resource management tools can also enhance and organise staffing allocation, minimising the need for short notice ‘over-time’ costs and maximising customer satisfaction by proposing slots that are more likely to be chosen.

Taking on-time delivery as a Key Performance Indicator (KPI), Bip recently helped a multinational agribusiness implement a predictive module that forecasts the KPI’s evolution and works out the root causes for late or early deliveries. The module’s predictions were impressive with 70% precision and resulted in an 11% improvement in the company’s punctuality and the identification of the 10 main root causes of unpunctual deliveries.

AI ultimately can be a vital tool for businesses to maintain good relationships with customers and trade partners. Around 96% of supply chain professionals are planning to use the technology going forward, according to Freightos.

Oversight and governance

Naturally, there have been some concerns about AI and the changes it will bring to the workplace, for instance, potentially taking jobs from humans or leading to dependency on the technology. However, it is important to be aware that AI can still make mistakes. AI is only so accurate as the data it’s fed and it can have inherent biases from flawed input or training data. Therefore, human oversight and governance remains a necessity.

Businesses should ensure they have a detailed AI strategy and governance framework to guide responsible implementation of the technology and guarantee compliance with wider regulation policy. This should incorporate company-specific AI goals, outline procedures for ensuring data is accurate and involve continuous human supervision, checking for error or biases.

Training employees on using AI productively is also important to promote efficiency and to free up their capacity to work on other ’value add’ tasks. Ultimately, AI can complement the capabilities of humans and allow for more effective logistics and supply chain management, however, still requires cautious management to maximise its potential.

[1]Orders which cannot be fulfilled due to a lack of inventory availability

Orders which cannot be fulfilled due to a lack of inventory availability