From Dynamic Updates to AI: How Tech Turbocharges Last-Mile Delivery

Last-mile delivery provider Jitsu has found a sweet spot of providing two to three delivery notifications to the customer per shipment: letting the customer know it’s on the way and when the truck is nearing the drop-off location.

“Customers love the information,” says Raj Ramanan, CEO of Jitsu. “And shippers love it because their support teams don’t have to answer the ‘where’s my order?’ questions.”

Dynamic Information Network

Delivery status notifications are only part of the technology revolutionizing last-mile delivery services.

Underlying many technological advances is the tectonic shift from a static, batch-processing model to a real-time dynamic network model. Instead of sending and receiving information once or twice a day, information must be available to all supply chain parties at the same time, all at once.

In a network environment, companies can see all the data through various internet-connected applications without having to move it from point to point.

“The network mimics what happens in the physical world, and because of the growth in the last-mile delivery network, shippers and their customers need information that is real time,” says Guru Rao, CEO of nuVizz.

Network-based technology enables a holistic overview rather than a siloed assortment of solutions, such as separate customer relationship management, or order and fleet management.

“Shippers want one platform where they can see all of their shipping activities and manage any breakdowns in the system,” Tejani says. “It’s not just about technology; it’s more about how businesses are evolving.”

Optimizing for Real

Optimization is a common buzzword in supply chain and logistics, but it means more than making operations slightly better. Tech is driving real routing and load management optimization, so more work gets done with fewer resources. Over time, it can result in fewer trucks on the road, reducing costs and improving sustainability.

Third-party logistics provider Hub Group is moving from standard slot-based routing to intelligent software that recognizes the best loading and routing for a truck carrying, say, a kayak and a refrigerator in a geographic area.

While fully utilizing the truck for the cargo is important, the types of services required add another layer of complexity.

“The delivery provider has to know if they are making a basic threshold delivery or a complex installation,” says Scott Robider, executive vice president for final mile at Hub Group. “The provider not only tries to factor the various pieces onto a truck but also tries to understand the time required for each of those services.”

AI to the Rescue

Once a route needs 15-20 stops, it becomes difficult to manage all the options. Artificial intelligence (AI) can take into account multiple variables—such as delivery windows at specific destinations and order cutoff times—to dynamically optimize on the go.

Because it’s fast and accurate, AI enables trucks and drivers to react on the fly to address changes, pick-up requests, or anything else that comes up during the route.

“The technology helps delivery providers figure out solutions so they don’t have to send another truck and driver. That’s a huge savings day in and day out,” notes Poberschin.

While 100% asset utilization is a goal, it is not realistic. Instead, providers can create a plan based on the 80% of demand that’s static and be prepared to handle 20% of the demand that’s dynamic, Rao suggests.

Final-mile delivery can take many forms—from a company’s fleet to parcel carriers, LTL networks, and gig economy workers. It’s critical for shippers to have the same visibility and transparency regardless of mode. The reality is, it can be difficult for a shipper to know what’s going on with an order in real time. But with the right technology solutions, it’s like the shipper is operating its own carrier fleet.

“Shippers can see where the driver is and where a shipment will be delayed,” says Khaled Naim, CEO of OnFleet. “They can even have the communications from the recipient routed to them directly.”

Consumer service expectations have spilled over into the business world, as people who are used to fast home deliveries from Amazon want the same speed at work. For example, food distributors now offer three-hour service to deliver restaurant supplies, when in the past a manager would send a waiter to a nearby store to replenish the napkin stock.

“Food distributors have enhanced their technology to better match up to customer systems and build out their own delivery network for the restaurants,” says Sameer Tejani, a director at global strategy consulting firm Stax.

Retailers and merchants explore last-mile options to manage capacity or enter new markets. Hub Group works with large retailers that want to offer daily delivery seven days a week instead of five, for example, or move to delivery six days a week in a low population geography.

Retailers want to compete on speed to market, so when a customer places an order it will be available when they want it. Some shippers offer free shipping in certain ZIP codes and increase marketing efforts to fill that capacity.

“A lot of shippers are starting to waive delivery fees, but they still have the burden of paying for it,” Robider says.

Ultimately, setting up a mattress or delivering a new garden swing may be a brand’s closest interaction with the customer. However, that critical piece of the relationship is out of the shipper’s hands.

“The technology projects the brand on the doorstep, and the carrier has to be a transparent extension of that brand to provide a closer customer experience,” Ramanan says.

Last-mile tech companies focus on making sure any driver can complete a delivery that makes the customer happy, taking into account details on which doorstep to leave a box or what times to avoid.

When he joined Jitsu about a year ago, Ramanan delivered a few packages to test the system. During one delivery, after he found the back alley spot to drop the box, and was taking a proof-of-delivery photo, the homeowner came out and picked up the package. The rules say no people are allowed in the proof photos, and the technology flagged the image. He immediately got a call from the service center notifying him of the error.

“The technology ensures there is professional-level delivery supervision at every doorstep,” Ramanan says.

When something goes wrong, it’s not the regional parcel carrier who takes the blame. Consumers are quick to call out the brand’s CEO on social media to fix the problems.

“Consumers don’t contact the call center or claims department; they go out on social media and tell the world about a delivery issue, so it’s important we get it right,” Robider says.

Reshaping the Landscape

Advances in machine learning and artificial intelligence are reshaping the last-mile landscape.

Jitsu is experimenting with multimodal AI that can incorporate photos and video as well as text information to analyze each delivery. The models are trained on thousands of successful and unsuccessful deliveries to spot problem areas.

The system can create an ever-tightening geofence around an address that gets repeated deliveries. It pinpoints the location of the porch or back stoop to ensure parcels are left in the right spot.

The system can intake operational changes and apply them to tasks moving forward. For example, Jitsu’s app was able to automatically reroute around the Francis Scott Key Bridge collapse in Baltimore. Once the system understood the route was no longer there across the body of water, it automatically reconfigured and repriced all the routes offered to drivers.

“This tech allows human-like reasoning at a distance and at a level of scale that previously wasn’t possible,” Ramanan says.

Improving efficiency by one tenth of one percent adds up over millions of stops. AI tools can help answer common questions from field personnel: Is this the correct address? What’s the gate code? AI can respond faster than a call center, especially when a driver is behind the wheel.

“That may seem trivial,” Ramanan says, “but every fraction adds up to be more competitive in the market.”

---

Pick-Ups Pick Up

Parcel pick-up locations—or out-of-home options—help shippers control costs and reduce the threat of porch piracy from the last mile.

Door-to-door shipments can get bogged down in traffic and leave customers’ purchases vulnerable to theft. Rising fuel and labor costs drive up the expense for each doorstep drop-off.

Overall, home delivery is the least efficient transportation option, according to Kearney research. Deliveries to a business usually have lower rates compared to a residence because carriers can complete multiple deliveries in one stop. In many cities, customers can choose from a network of convenient and secure pick-up locations.

Major pick-up outlets include Walgreens and Dollar General, which host UPS or FedEx access points.

Large retailers including Home Depot, Kohl’s, and Target offer pick-up and drop-off lockers onsite for customer convenience. Retailers see parcel services as a marketing tool, as customers tend to shop while they pick up a delivery at the store.

Walmart reports it lowered the cost of home deliveries by 20% in 2023 with an emphasis on store-fulfilled sales and parcel stations that stage products for home delivery by the company or independent drivers. Delivering more packages per route helps Walmart “densify the last mile,” said CFO John D. Rainey in an earnings call.

Out-of-home locations are more common in Europe, where bulk volumes surpass door deliveries. In the United States, lower delivery costs could entice more shoppers to accept delivery at alternate locations.

---

Fast Food, AI-Style

Few products are as perishable as a steaming burger and fries or a piping hot burrito. The delivery experience for a quick-service restaurant (QSR) can make or break their bottom line.

Botched deliveries not only irritate customers, but they also result in food waste; up to 15% of prepared food is tossed out due to delays and lack of temperature control.

With so many variables complicating on-the-fly adjustments, AI-enabled tech can help QSRs make better decisions. One such technology has helped QSRs cut delivery times by 35%.

Logistics management company LogiNext provides a solution that uses AI to automate QSR delivery operations. The technology leverages logistics techniques such as automated order assignment and allocation integrated with First-in-First-Out (FIFO) order assignment.

The AI intelligently assigns tasks to delivery agents based on availability, proximity, and capacity. The system takes the guesswork out of which driver can best handle an order. The FIFO program assigns the oldest orders to be delivered first, ensuring customers receive hot and fresh food and reducing the risk of food loss.

The AI-driven decision-making incorporates real-time data such as traffic and weather, along with order volumes, to support dynamic route planning for fast deliveries. Over time, QSRs can leverage predictive analytics to anticipate peak busy periods and adjust staffing and deliveries accordingly.

The LogiNext system allows QSRs to mix internal and external delivery carriers, enabling chains to adapt to varying demand levels and dynamically allocate resources, optimize their operations, and minimize delivery times, says Dhaval Thanki, executive vice president of LogiNext.

---