BLOG
Generative AI: Why smarter supply chains are here
5-MINUTE READ
October 5, 2023
It’s clear to me that the dramatic arrival of generative AI means supply chain reinvention at an unprecedented scale. The impact? Almost every role and function across the supply chain will be transformed.
It’s not a question of whether supply chain management will be transformed, but by how much. So, what will supply chain reinvention, powered by this new technology, look like in practice? And what should supply chain leaders do to prepare?
Let’s look at some facts
Accenture Research recently completed a study to estimate the impact of generative AI on roles across all company functions – including supply chain.
The study assessed 923 roles by determining which of their associated tasks required an intensive use of language. These roles (from O*NET), published by the US Department of Labor, have occupation-specific descriptors and cover almost the entire US economy. For each role, the database includes information on the mix of knowledge, skills and abilities required to complete it, as well as the activities and tasks that need to be performed.
In all, over 19,000 tasks were evaluated for the degree of communication, reasoning, and validation required to complete them. Once the language-dependent portion of each task was determined, it was then separated into three further components: (1) how much of it can be augmented by gen AI, (2) how much of it can be automated by generative AI, and (3) how much of it is best left to humans to execute. Appropriately, the research team used GPT-4 with the dataset to complete this assessment.
While the study is based on US-data, we believe that the findings can be applied to supply chain roles in other countries.
Here are a few key findings:
- Approximately 24% of all tasks evaluated in our study were determined to have a high potential for transformation with the use of generative AI. Transformation includes both augmentation, where generative AI works in conjunction with a human worker to enhance their ability and/or output, and automation, where generative AI can carry out the tasks instead of a human worker.
- Corporate functions where tasks have the highest potential for transformation by generative AI are IT/Technology (73%), Finance (70%) and Sales (67%); functions where tasks have lower potential for transformation are: HR (57%), Marketing (56%), Legal (46%) and Supply Chain (43%). This data reflects that while other functions have more language driven tasks, we believe that supply chain will still see dramatic benefits.
- Across the 43% of supply chain tasks with high potential for transformation, 29% of time will be impacted by automation; the remainder, 14%, will be impacted by augmentation, where generative AI can perform the tasks instead of a human worker.
When looking at how generative AI will transform work in supply chain-specific occupations (15 roles in total), we found that:
- Most of the supply chain occupations have a relatively high exposure to generative AI. In seven out of the 15 supply chain occupations, generative AI could affect more than half of all hours worked. (Refer to figure below.)
- Slightly more than 30% of people’s time is currently spent on tasks that could be automated by generative AI. Uppermost among these are tasks in production, planning and expediting.
- Almost 17% of people’s time is spent on tasks that could be augmented by generative AI. Taken together, automation and augmentation to this degree looks set to drive average productivity savings of almost 20%.
How Generative AI will transform work in supply chain specific occupations.
Work time distribution by occupation and potential Large Language Model (LLM) impact. (Listed in order by employment levels in the US in 2022.)
Note: Estimates are based on Human+Machine identification of work tasks exposure to impact of generative AI.
Source: Accenture Research based on US BLS May 2023 and O*Net.
How will generative AI support the supply chain workforce?
Considering that supply chain roles have a 43% potential for transformation, let me run through some of the principal ways in which generative AI supports people – and the potential impacts it could have:
- Advising: Generative AI puts new kinds of hyper-personalized intelligence into human hands. Category managers, supply chain planners, field engineers and plant workers should all benefit. In-scope areas include advising on vendor selection within specified parameters, providing step-by-step guidance on asset management processes, and assisting with real-time supply chain alerts.
- Creating: As a creative partner to product engineers, retailers and materials planners, generative AI should reveal innovative new ways to engage audiences and revolutionize areas like product design. A couple of examples? Solving for sustainable product packaging or creating new materials and parts designs.
- Coding: Data scientists and analysts working in supply chain operations will use generative AI to dramatically accelerate conversion of programming languages and become proficient in new programming techniques. Productivity should surge as a result.
- Automating: With its sophisticated understanding of historical context, next best actions, summarization and predictive intelligence, generative AI will take business process automation to a completely new level across front- and back-office operations. Use-cases in the supply chain include using next-gen chatbots to enable hyper-personalized customer service interactions and transforming contract management by automatically drafting new agreements based on historical data.
- Protecting: We’ll see generative AI being used as a vigilant co-worker to proactively identify risks – supporting sustainability and compliance managers and safeguarding against fraud.
Just like AI up to now, I’m convinced we’ll see the most transformational results from generative AI when it is scaled across enterprise supply chains. One example? Think about the benefits for supply chain resiliency and agility.
In one real-world use-case, generative AI virtual assistants are being used to help supply chain managers secure new visibility into and control over supply chain disruption. Having alerted managers to possible disruptions to shipments (using AI-driven news reports), the virtual assistants can identify which suppliers are most likely to be impacted and prepare emails for rapid and seamless transmission to them (including suggested questions to help managers better understand planned mitigation initiatives.)
Or take another example: the impact on manufacturing and technical operations. Today, technicians gather supporting documents for key assets and systems from a fragmented knowledge base and use this information to support critical incident solving, maintenance and troubleshooting.
Using generative AI, they’ll have access to a single interface providing a new, integrated knowledge base and rapid cognitive insights. This will guide troubleshooting and root-cause analysis, provide key transactional information and trigger actions, and facilitate shift handover management. Our research finds that 25% of total workhours in Installation, Maintenance and Repair will be affected.
Getting started: adopting generative AI into supply chain operations
To get the adoption journey underway, CSCOs and CPOs should identify the potential of AI (with a focus on generative AI) in their existing processes. For those processes with higher potential, they should evaluate the automation impact as well as the new value generated.
This “2 by 2” matrix will show where they need to prioritize their efforts from the outset. In other words, to scale generative AI faster and move to a new performance frontier, supply chain leaders should take a “think big, start small” approach.
We’ve identified a number of imperatives that should underpin the adoption and scaling of generative AI. I’ve summarized them here, but it’s well worth diving into the PoV to take a closer look:
- Embrace a business-driven mindset toward generative AI adoption; it’s the key to define and successfully deliver on the business case.
- Build talent pipelines with the data science skills to take foundation models, adapt them to business needs and integrate them into your applications.
- Data integrity is vital, so ensure a strategic and disciplined approach to acquiring, growing, refining, safeguarding and deploying data.
- Make sure you have a sustainable technology foundation to meet the high compute demands of generative AI, while closely monitoring cost and energy consumption.
- Leverage the substantial investments in generative AI by cloud hyperscalers, big-tech players and startups so you don’t need to go it alone.
- Focus on being responsible by design – that means moving from a reactive compliance strategy to proactive development of mature Responsible AI capabilities.
Generative AI marks a step-change in how and where organizations, and their supply chains, use AI. The opportunities are, frankly, enormous. And the time to start testing them out is now.
I’d love to hear your thoughts, so please get in touch. Thanks for reading.