the ConversAItion: Season 6 Episode 35

How UPS Leverages AI to Level Up Logistics

UPS delivers nearly 25 million packages and documents every day. While the process seems near-effortless, AI is hard at work behind the scenes to ensure the reliability consumers have come to expect. Today on The ConversAItion, Laura Patel, Principal Data Scientist within UPS’ Advanced Technology Group, walks us through how the scientific method can inform data science, the role of AI in a package lifecycle and why data will be essential for the success of logistics companies going forward—including building a more sustainable future.
Listen on Apple Podcasts badge
“AI and emerging technologies are key to understanding our complex network. We need a vast amount of data collected at every stage of the process—and can use those data and AI models to track packages throughout the facilities, increase vehicle utilization and optimize our last mile routing.”
Laura Patel

About Laura Patel

Dr. Laura Patel is Principal Data Scientist at UPS, where she oversees a team that uses AI and advanced data analytics to process package volume as efficiently and reliably as possible. She is an expert physicist, and has published numerous research papers on nuclear and particle physics. Laura earned her PhD, Master’s and Bachelor’s in Physics from Georgia State University, and currently resides in Alpharetta, Georgia. She can be found on LinkedIn here.

Short on time? Here are 5 quick takeaways:

  1. Statistical analysis can be just as relevant at UPS as it is in a physics lab. 

    Prior to joining UPS, Laura garnered an impressive background in academia, earning a Master’s degree in physics and a PhD in nuclear and particle physics. In fact, it was this expertise that empowered her to make the transition to tech and data science. Why? Much of the statistical analysis involved in physics can be transferred over to the business world; in fact, she believes that’s what data science is at heart. Moreover, Laura’s expertise in the scientific method, including best practices for designing experiments, proved to be essential at a fast-paced, cutting-edge company like UPS.

  2. AI helps ensure that UPS packages get to their destinations as efficiently as possible. 

    The lifecycle of a typical UPS package is heavily dictated by AI. When a package enters the UPS network, whether through driver pickup or customer dropoff, it travels to a variety of sorting hubs based on its destination. After arriving at its final hub, the package is loaded onto a hallmark UPS brown truck—one of almost a hundred thousand—and delivered to the right customer.

    This multi-step process is highly complex and spans multiple geographical locations. UPS leverages data and AI models to make it simple; the technology helps track packages in each sorting facility, optimizing vehicle use and routing to ensure that a package gets delivered in the most efficient manner possible. The more data UPS has at its disposal, the more its algorithms can learn—and the more its processes can be improved.

  3. UPS uses AI models to analyze billions of data points and, in turn, optimize the customer experience.

    UPS’ Advanced Technology Group is on a mission to improve the overall customer experience with emerging technology. Laura’s analytics research subgroup leverages data and AI to make more robust, informed business decisions that power UPS’ global smart logistics network, which converges all UPS data streams in one place to connect packages with customers. This data-driven decision-making process is powered by billions of data points collected by UPS—over one billion per day—coupled with the latest in machine learning, optimization, simulation and visualization. These AI models ultimately help with customer experience objectives, like getting packages delivered more quickly and providing transparency into ETAs—even amidst supply chain disruptions.

  4. Amidst pandemic-induced shifts in delivery patterns and demand, UPS leaned on AI to stay nimble. 

    During the pandemic, consumers flocked to online shopping to buy almost everything, from groceries and prescriptions to clothing, rather than visiting brick-and-mortar stores. As a result, UPS saw an increase in shipping volume, making the last mile delivery routings—getting packages from their final sorting hub to the customer’s doorstep—all the more complex. UPS’ existing algorithm for last mile routing, called “Orion,” allowed UPS to optimize driver routes based on volume and traffic, thereby reducing millions of miles from routes. 

    Despite the multitude of shifts taking place in the supply chain during the pandemic, UPS was able to not only meet customer expectations, but also be a part of the solution. Today, the company uses revolutionary radio frequency identification (RFID) technology for vaccine tracking, and has already facilitated the delivery of over one billion vaccines. 

  5. AI is poised to shape the future of logistics in customer experience, sustainability and beyond. 

    According to Laura, the volume of data worldwide has been increasing at a nonlinear pace. With a growing amount of available data points every day, robust AI models will ensure that logistics companies can meet customer needs, make improvements across the supply chain network and address any global shifts, whether due to natural disasters, pandemics or other unforeseen shocks to the supply chain. 

    Within the broader logistics industry, Laura believes that AI will play a critical role in sustainability efforts, as there will be an increased focus on environmental impact with lower emission vehicles, alternative fuels and more. UPS is already leveraging AI to achieve its mission of having 40% of ground operations using alternative fuels by 2025, and being completely carbon neutral by 2050.

Read the transcript

TRANSCRIPT

EPISODE 35: Laura Patel

Jim Freeze Hi! And welcome. I’m Jim Freeze, and this is The ConversAItion, a podcast airing viewpoints on the impact of artificial intelligence on business and society. 

[UPBEAT MUSIC]

On today’s episode of The ConversAItion, I sit down with Laura Patel, Principal Data Scientist at UPS. Since the pandemic hit, most of us have relied on UPS more than ever before to send and receive packages, but there’s actually a lot more going on behind the scenes. Laura’s here to give us an inside look; we’ll discuss AI’s role in helping UPS deliver nearly 25 million packages a day, how supply chain shortages accelerated AI innovation and forthcoming opportunities for AI in the logistics space.

Laura, thanks so much for joining us. We’re thrilled to have you!

Laura Patel Thank you for having me.

Jim Freeze So, you’re very accomplished academically, multiple degrees including a PhD in physics. Can you tell us a little bit about how and why you made the pivot to AI, and how your work in physics informs what you do?

Laura Patel Yeah. I’ve always been fascinated by how math can be used to study the natural world. This is one reason I’ve decided to pursue graduate work in physics. So, I started with solid state physics for my master’s degree, and then changed specialties while remaining in physics, to get my PhD in nuclear and particle physics. So what does that have to do with AI? Well, there’s a lot of statistical analysis involved in physics processes that are used specifically to understand the fundamental properties of physics, but a lot of that can be transitioned over into the business world, especially with AI and tech.

I decided to make the shift to data science after graduating. It was a newer field at the time with exciting and challenging opportunities to apply the scientific method to solve business problems. To me, that’s really what data science is at heart. Skills that have helped along the way, not necessarily the laws of physics, but other areas that you learn. How to approach challenges, how to design and set up experiments, communication skills. There have been areas where some physics has helped, but for the most part, I think anything that’s challenging and able to learn from is just a great way to get involved.

Jim Freeze I love that answer. And just for whatever it’s worth, I have a master’s degree in mathematics. I have a law degree. I passed the bar, and I’m a marketing executive, so I loved your answer. So at UPS, I understand your team advances data analytics to process package volume as efficiently and reliably as possible. Can you walk us through your role in your team’s mission in that regard?

Laura Patel Absolutely. I work in the advanced technology group, which brings together emerging technology to serve customer needs and improve the customer experience. Within that group, we have an analytics research subgroup, where I’m the principal data scientist. So in my group, we focus on how to use data and AI to make more robust business decisions, so this idea of data driven decision making. And we do that to power our global smart logistics network, to be able to seamlessly connect packages and customers. UPS collects over one billion data points a day, and the analytics research group’s able to use our special skills in machine learning, optimization, simulation, and visualization to make this vast amount of data into AI models to help support UPS initiatives.

Jim Freeze That’s interesting. UPS delivers 25 million packages every day, which is pretty incredible when you think about it. What’s the tip of the life cycle of a package, and how is AI there to help ensure that it gets where it’s supposed to be?

Laura Patel So to start with, a package will enter our network. This can either be through driver pickup or customer drop off. The package then travels to an origin sortation facility, one of over 1000 we have domestically. And then a package is transported to another sortation hub by either plane, train or truck. This can happen multiple times depending on the lane that the package is moving in. And then the package will arrive at a destination sortation facility where it’s loaded onto the delivery vehicle.

So the package cars, which I think is what a lot of people associate with UPS, our brown cars—your package will go on one of almost a hundred thousand vehicles that will arrive at the customer house or desired location. Now, AI and emerging technologies are keys to being able to understand such a complex network, so for these efforts, we need a vast amount of data collected at every stage of the process. As far as AI along the path that the package takes, we can use the data and AI models to track packages throughout the facilities, to increase vehicle utilization, and to optimize our last mile routing.

Jim Freeze It’s a hundred thousand vehicles. That’s pretty remarkable, and you see them everywhere. Over the course of the past two years since the pandemic hit, the whole notion of supply chain has become a much more talked about and common thing in everyday life. The shortages, supply chain shortages have been top of mind for lots of folks. And obviously, the pandemic has made it just a critical daily issue that people discuss. I’m curious as to how the shortages and delays that have been experienced over the course of the past couple years have impacted your AI strategy.

Laura Patel For the most part, delays are occurring farther upstream than our small package network. Once we receive packages into the network, we’re able to deliver with a high service performance. We have learned a lot during the pandemic though, and we’re able to use these learnings to enhance our tools through continuous improvements. So for example, UPS, we have some revolutionary RFID technology for vaccine tracking that’s allowed us to deliver more than one billion vaccines.

Jim Freeze Wow. Talk about something that’s impactful. A billion vaccines have been delivered through UPS.

Laura Patel Yes.

Jim Freeze That’s pretty remarkable. Were there any other trends you noticed over the course of the past couple years as a result of the pandemic?

Laura Patel We have seen an increase in our volume in the B2C area, a lot you had mentioned earlier with online shopping. And that has presented some unique challenges and opportunities within the AI space, especially for our last mile delivery routings. We have a very robust, flexible system, Orion, which is our last mile routing algorithm. We use that to optimize driver routes, and we’re able to cut millions of miles off of our routes per year by doing this optimization. And so with changes in our last mile delivery due to this B2C transition, we’ve been able to meet customer needs by addressing the challenges that AI can present.

Jim Freeze Interesting, interesting. One of the things we always love to do on The Conversation is kind of ask our guests to be prognosticators, so I’ll ask you to talk a little bit about the role of data and AI. How is it going to evolve in the coming years in logistics in particular?

Laura Patel Data has been increasing at a nonlinear pace, the amount that’s collected just in general throughout the world. And this data within supply chains, they can be complicated, and AI can be leveraged through emerging technologies to meet customer needs. So as I mentioned earlier, the data we need throughout the entire supply chain each day, we’re collecting more and more data, and that can help feed into building more robust AI models and allow UPS and other logistic companies, and companies in general, to unlock future operations. This can lead to continuous improvements across the network.

At UPS we’re dedicated to our global smart logistic network, which is the ability to seamlessly connect packages and customers through AI, and having a very flexible adaptable network to address any changes in customer behaviors or natural disasters or any sort of unforeseen changes in the supply chain. 

For the logistics industry as a whole, I think there’s going to be an increased focus on environmental sustainability as we move forward, with lower mission vehicles, e-vehicles, and alternative fuels. And UPS has set sustainability goals for 2025, and to eventually by 2050, be carbon neutral. I really believe that AI is necessary to reach these goals.

Jim Freeze That’s interesting. Will that probably entail by, of course it will, a completely different fleet for delivery? I mean, everything’s got to go EV.

Laura Patel Yeah, we’re in the process of looking into ways using data and AI to be able to reach carbon neutrality. And also our goal for 2025 is 40% of ground operations with alternative fuels. So it’s going to need to be a different combination that AI can hopefully help to optimize, in order to reach these goals.

Jim Freeze Fantastic. Laura, this has been great. I’ve learned a lot, and I know our listeners will learn a lot. Really appreciate you being on the podcast. Thank you so much.

Laura Patel Thank you.

[OUTRO MUSIC]

That’s a wrap for this episode of The ConversAItion. We’ll be back next time to chat with Ella [Hill-all], Vice President of Data Science and Engineering at Shopify. She’ll share how Shopify applies AI to optimize transactions, and where she sees the role of AI headed in e-commerce in the coming years. You won’t want to miss it!

This episode of The ConversAItion podcast was produced by Interactions, a conversational AI company. I’m Jim Freeze, signing off, and we’ll see you next time.

[UPBEAT MUSIC]

+++

 

Check out more episodes of The ConversAItion.