How supply chain management can be transformed by generative AI

Artificial intelligence (AI) has become an increasingly important topic these past years. This brought about a transformative shift in various domains, with generative AI standing at the forefront of this revolution.

Generative AI (GAI) is a machine-learning model that is trained to create new data. We all know, for example, OpenAI’s GPT-4 and DALL-E ; these models are capable, in some ways, of mimicking human thinking and creativity (1).

More specifically, generative AI is emerging as a transformative force in supply chain management. How does it impact this field ? And how can we put GAI to work in managing supply chains, despite its limitations ?

Sachin Kamble, Professor of Strategy (Operations and Supply Chain Management) at EDHEC Business School, explores in this article what exactly is Generative AI, and how it can benefit supply chain management (2).

First and foremost : what is Generative AI?

Generative Artificial Intelligence is an advancement in the capabilities offered by artificial intelligence-based technologies. Unlike traditional AI, which is mainly focused on analyzing data and making predictions, generative AI is designed to create new content. For example, it can produce images, music, text and complex designs, demonstrating remarkable capabilities in imitating human creativity.

While AI is focused on handling specific tasks, GAI develops cognitive systems capable of understanding, learning, and applying intelligence across different topics. Its applications are varied (3), ranging from accelerating scientific research to enhancing creative processes in specific industries.

The key technologies behind GAI include deep learning and natural language processing (NLP) algorithms. Deep learning uses algorithms to simulate the human brain process. It can perform tasks ranging from recognizing images and speech to translating languages. NLP enables computers to understand (and communicate with) human language.

One of the most famous GAI models is GPT-4, which produces lifelike text. Chances are, you’re already familiar with it. It allows the user to submit a request or a question, prompting the AI to generate new content. Other notable GAI tools include :

• Scribe, which is an AI writing assistant used for content creation;
• AlphaCode and GitHub Copilot, which are coding assistants that provide developers with code snippets and context-aware suggestions for coding tasks;
• Dall-E2, which creates visual art and illustrations;
• Duet AI, which assists artists in creating music compositions.

How can GAI be used in supply chain management?

Generative artificial intelligence can be used in many areas, including supply chain management. Actually, it has the power to make the supply chain technology much smarter in terms of efficiency and productivity.

Companies have already been deploying AI in supply chains for demand planning and procurement, while exploring its use in other areas (inventory, logistics, demand planning, etc). According to an EY study (4), 40% of supply chain organizations are currently investing in Generative AI.

Here are some examples of GAI use in supply chain management:

Plan (Processes that aim at balancing the aggregate demand and supply)

GAI can assist in demand forecasting (by using past data), inventory planning, production planning and risk mitigation.

Source (Processes that procure goods and services)

GAI can be a powerful tool in supplier management, sourcing and contract management.

Make (Processes that transform the raw material into finished goods)

GAI can generate and evaluate designs, and create predictive maintenance schedules.

Deliver (Processes that put the finished goods and services to the market)

GAI can help with global network optimization and logistic network designs.

Return (Processes focused on product returns)

GAI can be used to predict how many returns there will be and find the best ways to deal with them.

Apart from the above uses, GAI can be used to reach sustainable objectives. For instance, training generative AI models on supply chain performance data on sustainable parameters like material use, pollution levels, recycling, and renewable materials can lead to making the processes more sustainable. This is important, since more than 90% of an organization's greenhouse gas emissions and 50% to 70% of operating costs are attributable to supply chains (5)

Understanding the benefits of GAI in supply chain management

As we have seen, companies can integrate GAI into these five building blocks of supply chain operations : plan, source, make, deliver, return. This implies a vast potential benefit for supply chain practitioners.  

So, in concrete terms, how can GAI build efficient and resilient supply chains ? Here is a list of its most important benefits:

• Cost reduction (because repetitive tasks, such as tracking and documenting inventory, can be completed with greater accuracy and less labor).
• Improved safety (thanks to AI systems being able to monitor work environments and detect conditions that threaten the safety of workers).
• Increased efficiency and optimized operations.
• Improved customer satisfaction.
• Enhanced decision-making.
• Improved sustainability.

What are the limitations and challenges of GAI ?

Using GAI to manage supply chains can’t be done overnight. While the technology has significant potential, it also has its share of challenges and limitations.

Unfortunately, these challenges can limit the deployment of GAI in supply chain management and act as potential barriers. Here are some of them :

Training costs

As with any technology, implementing AI and integrating it into production environments requires (costly) training.

Limited training data

The performance of GAI depends on the availability of extensive training data, which can be challenging for many supply chains as the data resides with different stakeholders.

Data diversity

Obviously, GAI requires standardized and high-quality data to make accurate predictions. However, in most supply chains, the data is not standard due to the use of different systems, equipment, and data formats. Furthermore, the data could be diverse and fragmented, with inconsistent quality levels. This inconsistency can lead to distorted results.

Deployment and scalability

Deploying GAI in the supply chain management can lead to substantial financial and infrastructural challenges. Indeed, integrating GAI with the existing technologies requires a robust and scalable IT infrastructure with the capability of handling complex data processing.

Real-time adaptation and dynamic environments

As we have seen, GAI is mainly built on past data. Therefore, it can face challenges in quickly adapting to unforeseen events (such as natural disasters, transportation delays…), which can result  in poor decision-making.  

Privacy and ethical considerations

Last but not least, developing GAI requires to consider ethical and legal concerns. We know that GAIs are based on volumes of data, which can pose issues with data security, privacy, and the ethical use of data. These situations demand that the supply chains strictly follow the rules, such as the General Data Protection Regulation (GDPR).

To conclude

Over the years, GAI has become a disruptive technology within the context of supply chain management, making all the sub-processes like plan, source, make, deliver and return smoother. It is expected to positively impact the supply chain management, fostering increased productivity, safety, sustainability and creativity.

However, deploying GAI in supply chains comes with its own challenges : training costs, ethical considerations, lack of data diversity…. It is essential to acknowledge these potential issues, so that it becomes easier to navigate them effectively.

In the near future, supply chains will then be able to rely on GAI to optimize their operations, and thrive in an increasingly competitive marketplace.

Photo by Ashley via Unsplash

References

(1) AI Mimics Human Creativity, But Also Boosts It (2023), Forbes - https://www.forbes.com/sites/joemckendrick/2023/07/16/ai-may-or-may-not-be-able-to-mimic-human-creativity-but-certainly-can-boost-it/

(2) This EDHEC Vox paper is based on the research conducted by Prof. Kamble on GenAI and suuply chain, including a selection here of 3 key ones:

• How to use no-code artificial intelligence to predict and minimize the inventory distortions for resilient supply chains (2024), International Journal of Production Research - https://doi.org/10.1080/00207543.2023.2166139
• Adoption of Artificial Intelligence and Cutting-Edge Technologies for Production System Sustainability: A Moderator-Mediation Analysis (2022), Information Systems Frontiers - https://doi.org/10.1007/s10796-022-10317-x
• Artificial intelligence-driven innovation for enhancing supply chain resilience and performance under the effect of supply chain dynamism: an empirical investigation (2021), Annals of Operations Research - https://doi.org/10.1007/s10479-021-03956-x

(3) The Generative AI Dossier - A selection of high-impact use cases across six major industries (2023) Deloitte AI Institute - https://www2.deloitte.com/us/en/pages/consulting/articles/gen-ai-use-cases.html

(4) How sustainable supply chains are driving business transformation (2022) EY - https://www.ey.com/en_gl/insights/supply-chain/supply-chain-sustainability-2022

(5) Supply chain guidance - information for organizations interested in reducing their supply chain emissions (2022), US Environmental Protection Agency - https://www.epa.gov/climateleadership/supply-chain-guidance