Start small, scale up: How to integrate AI into automotive supply chains

By: Kim Bedford

Achieving end-to-end supply chain visibility and management has always posed a challenge for Tier 1 automotive companies. With the help of AI, however, the automotive industry is beginning to realize that change is possible. In fact, Gartner^® has reported that “top supply chain organizations are using AI to optimize processes at more than twice the rate of low performing peers.”¹

Yet, in conversations with industry leaders, I’ve observed that many still grapple with the fundamental hows: how to get started with AI and then how to scale up. Some feel overwhelmed — even frozen — by the sheer scope of possibility, while others risk overextending their resources and budgets by rushing to deploy AI across their entire supply chain.

For teams that might still be stuck on the first how — how to get started — here are three areas where I believe AI will have the biggest impact on our industry in the year to come.

Streamline processes and customer service operations

Exploring AI applications for operations might start with a prioritization matrix. What are the main areas for improvement? Which weak spots present the highest risk to business continuity or your brand?

One of our customers, a leading global car rental company, was recently looking for support in managing its supply chain. An area in which they wanted to focus these efforts was vehicle availability, in order to elevate their offerings and service. In particular, they were eager to enhance their vehicle traceability, and their data collection methods related to fuel efficiency, damages and other key metrics.

One application of AI may be to help sync and streamline maintenance schedules and fleet rotations across rental locations. To target these goals, the AI tool this customer seeks out should be able to provide predictive insights to optimize spare parts inventories, ensuring the right parts are available at the right time. Another option may be to apply AI to demand planning, enabling more dynamic response to demand spikes and drops.

Improve compliance and security

Another significant challenge our customer faced was the lack of end-to-end visibility in their supply chain. This gap created many pain points, including the responsibilities of maintaining compliance with regulatory bodies and ensuring robust data security, especially concerning renter data.

How, then, might this customer use AI to improve existing compliance and security processes? And how could AI help them adapt more agilely to the evolving regulatory landscape?

Regulations such as the General Data Protection Regulation (GDPR) in Europe, the California Consumer Privacy Act (CCPA) and other sector-specific laws mandate increasingly stringent data privacy and security controls. Meanwhile, new or evolving technological and safety requirements put further demands on supply chains and production processes.

To stay on top of these regulatory and security concerns, our customer — and teams facing similar challenges — might consider leveraging AI to provide continuous monitoring, automated compliance checks, comprehensive risk assessments and robust incident response capabilities.

This application of AI can in turn help teams to enhance their security posture, meet regulatory requirements, build trust with both customers and stakeholders and, ultimately, boost resiliency across supply chains.

Enhance data analytics and integration

Making sure that the successes of your initial AI use cases are measured and tracked will be vital to hacking that second how: how to scale up.

This customer might therefore consider integrating an Al tool that consolidates supply chain data into a centralized dashboard. This consolidation can provide teams with a powerful compass for guiding any upscaling through increased visibility and data-driven insights into the performance of AI use cases.

For example, after our customer is able to prove the worth of their AI within inventory management and demand planning, they might expand their investment to explore additional ways in which to reduce vehicle downtime and enhance fleet availability.

After all, the goal of these first forays into AI applications within the supply chain is not just to establish proof of concept for deployment. It is also to identify areas within the supply chain where Al can be most effectively implemented in the future, ensuring that further investments are both strategic and customized to meet specific operational needs.

Consider a modular approach

One more point to consider: there is a crucial difference between scaling up and scaling up sustainably.

Before biting off more than they can chew, I encourage all customers I work with to consider a modular approach.

What does this mean exactly?

1. Starting with small, manageable projects that leave room for iteration and growth
2. Choosing projects that address recent and specific challenges, such as constrained supply, disruptive events or blockages, and the optimization of specific processes
3. Limiting the scope of projects to areas in which your team has a data strategy in place or that produces data that is accurate, consistent and complete

By stress testing your Al applications within these guardrails, your team will be better equipped to mitigate the possible risks and compliance concerns involved with trialing this technology — such as potential biases or quality issues — while also allowing for simple validation and clear and measurable results. This, in turn, will help your team to demonstrate the value of these AI applications to both internal and external stakeholders.

Kim Bedford
is Vice President of Automotive, Electronics, Manufacturing for Kyndryl US

¹ Gartner Says Top Supply Chain Organizations are Using AI to Optimize Processes at More Than Twice the Rate of Low Performing Peers, Gartner Press Release, February 2024. GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved.