Postal Applications of A.I.

Postal Applications of A.I.

Artificial intelligence is pretty much everywhere these days. It's shaping how we live, how we work, and even how our mail gets delivered. The postal and shipping industries are experiencing big changes. But what does AI actually look like behind the scenes in the postal world?

Hi everyone, I'm Theresa Klein, a research analyst with the US Postal Service Office of Inspector General. I'm joined by Rick Bauer, a research specialist here.

Hi, Teresa.

Rick, you're one of the primary authors behind our new report, Postal Applications of Artificial Intelligence. Before we dig into the details. Take us back to the beginning. What were the big questions you were trying to answer?

Yeah, sure. So in writing and researching this report, we were seeking to better understand AI's impact on the postal and logistics industries, as well as the implications for the future. So what we found was that all posts that we spoke with are experimenting with AI in some form, but they're all in different stages of implementation. But they do share common goals.

And those include improving efficiency, reducing costs, and improving the overall customer experience.

I have to say, when I think of the post office, I don't think AI is necessarily the first thing that comes to mind. Where is this tech actually being used behind the scenes by posts?

Well, we found that that there's quite a bit of piloting and experimenting going on in the industry, but we also identified quite a few established use cases, and those are in the areas of mail and parcel processing as well as delivery operations, customer service and back office functions.

For the most part, usage of AI is aimed at solving specific problems rather than trying to automate or redesign processes across the entire organization. However, we also found the industry moving toward more advanced applications such as real time dynamic routing, physical AI, computer vision, and interacting with customers via agenda AI,

Bringing it home to the Postal service. How are we putting AI to work right now?

Well, the Postal Service has always been an early adopter of AI. They've been using optical character recognition and chatbots for years now, and they continue to evaluate and introduce new applications across their operations. Currently, their primary focus is in mail and parcel processing and delivery operations.

So in mail and parcel processing, they're using computer vision to correct addresses in real time as well as to identify counterfeit postage. They're also testing a parcel robotic sorting system that can handle 3 to 4000 packages a day.

In delivery operations, the Postal Service is using AI to develop time and cost effective delivery routes, as well as to boost the reliability and predictability of their delivery services.

So what's on the horizon for the Postal service when it comes to AI? What do you think we should expect to see in the coming years?

Well, we see the Postal service moving towards more integrated systems.

They're beginning to use AI to to solve problems and coordinate processes. Areas of particular promise include network efficiency, customer engagement, revenue protection, and workforce productivity.

So to really make this future a reality, what actually needs to happen? What do you think are like the key ingredients to success?

So we found that those organizations that have a successful record of AI implementation have a number of things in common, and those include a culture of innovation, a well-defined strategic focus.

They tend to pursue only those applications that present a positive return on investment, and they also have access to the funding necessary to deploy AI. We also found that in order to be successful, it's important that organizations have the necessary technical capability, and that includes both human and computing resources. And finally, that they have the necessary data, which is both readily available and of sufficient quality to be of use in AI applications.

And finally, it's important that they comply with their own governance policies, and that includes both data security and ethics.

All right. So tell me the truth, Rick. Am I going to see little delivery robots at my door? Like, how far are we from having a fully autonomous post office?

Yeah, as cool as that might be, that's not going to happen any time in the foreseeable future. What we did identify was a three stage maturity curve for for AI in the postal and logistics industry.

Stage one is tactical support, and that's where AI is used as a tool to to both support and assist human workers. Stage two is value chain integration. And that is where AI connects and also coordinates data across different parts of the post operations. And stage three is system orchestration. And that's where AI autonomously makes decisions and triggers actions.

The postal service and the other posts we spoke with are somewhere between stages one and two.

The Postal Service's future competitiveness will largely depend upon their ability to move from their current isolated AI pilots toward more integrated, transformational systems.

Well, thanks for joining me, Rick, to talk about all things AI in the postal sector. For our listeners, if you're ready to check out the full white paper, you can find this one and all of our white papers at www.uspsoig.gov. 

Thanks for listening.