Last-mile logistics is more complicated than first- or middle-mile, and it's changing all the time, says Akash Agarwal, co-founder and chief business officer at Beans.ai.

Agarwal sums up the difficulty of last-mile delivery like this: “It’s not one-to-one or one-to-few, it's one-to-many. You're sending a truck to several addresses.” And that vehicle’s journey encompasses many actions that must take place if time is to be saved, money made, and customer satisfaction guaranteed.

As he sees it, “You have to create an optimized sequence to hit those addresses, and addresses themselves are very complex entities.” Insufficient information on where to park the delivery vehicle, where to enter the building, what floor to deliver to — these are among many challenges that make last-mile delivery either inefficient or even impossible, Agarwal says.

Multiply that by the ever-growing number of addresses a driver must deliver to, and you can see the enormity of the challenge, he says. Meeting promised delivery windows rests on having adequate address information.

Shippers, carriers and customers are the three partners in this intricate network, and each has its unique demands and requirements. The customer’s expectations are quite well known by now: He or she wants to know where their order is in the delivery chain and when it is to arrive. The shipper wants the same information because continued sales rest on meeting the customer’s demands. And a continuing relationship with the carrier rests on making both the shipper and the customer happy.

Agarwal says complete, verifiable information on the delivery address is key from the get-go. Second, a carrier must have a plan: how many trucks to deploy and in what sequence are they to “hit” delivery points. “So maps, planning and route optimization are the three core pillars.”