Delivering the Future LASTMILE SOLUTIONS:
deliveries, he says. Walker estimates the legislative foundation and technology infrastructure for scaled deployment will mature within 12 to 18 months. Building compliant commercial eets that meet rigorous U.S. safety standards will require additional time beyond that. Delivery Robots Innovation is also occurring on the ground. The delivery robots developed by Starship Technologies have made more than 9 million deliveries around the world. To identify and navigate around obstacles, the robots are equipped with ultrasonic sensors, neural networks, 12 cameras, and other tools. The “bubble of awareness” around them enables them to identify objects, like pedestrians, so they can stop before running into them, says Niclas Schwarz, business development manager. Getting a eet of robots up and running generally takes one to two months. Starship works with the organization to map out the routes the robots will travel. Some cities require permits before the robots can travel on sidewalks. The robots are able to cross streets and can travel up to about 20 kilometers, or about 12.4 miles, Schwarz says. They tend to be most cost-ef cient at between three and four kilometers.
FarEye’s order-to-door platform helps logistics and retail brands simplify last-mile complexity, ensure on-time delivery, and manage shipments with transparency and control.
beyond about 2.5 miles, says Ali Kashani, Serve’s president and CEO. Serve robots are on the ground in multiple U.S. cities and the company was on track to deploy 2,000 robots by the end of 2025. Route Optimization Software Along with hardware, the software that powers delivery vehicles is advancing. The platform from NextBillion.ai leverages AI, algorithms, and real-time data, like network demand and delivery policies, to optimize the last-mile delivery process, Bubna says. As real- world conditions change, the system can adjust and re-optimize routes, and can alert the dispatch and operations teams. Customers using the solution report 20 to 30% drops in delivery costs, as well as eet
at checkout, such as same-day or next- day delivery, based on real-time carrier capacity and business rules. The routing engine uses optimization heuristics and machine learning to create delivery plans that consider multiple factors, including traf c patterns, driver schedules, and service agreements. This sort of delivery intelligence has become critical, particularly in the United States, where “delivery ecosystems involve complex product categories, multi- carrier networks, and signi cant service variability,” Kumar says. Smart Electric Vehicles IndiGO Technologies, which hails from the mobility labs at MIT, is working to improve the technology that powers delivery trucks.
Starship’s robots are delivering groceries in about six European
countries. They can handle about 40 liters of volume, or the rough equivalent of two or three grocery bags of items. While it’s dif cult to pinpoint how much less expensive delivery robots are than couriers, most clients see an expense reduction, Schwarz says. The costs to deploy delivery robots are dropping by 20 to 30% annually, he says. In the future, delivery robots may link with drones. In late 2024, Serve Robotics, an Uber spinout, announced a partnership with Wing, the drone subsidiary of Alphabet, in which a Serve robot would automatically hand a package to the drone. This capability would be useful for deliveries that extend
utilization improvements of 15 to 18%, Bubna says. They also see signi cant improvements in the accuracy of estimated arrival times. FarEye offers an AI-powered last-mile technology platform that integrates with order management, warehouse management, and ERP systems to ingest order and inventory data. It then presents live delivery options
IndiGO Technologies’ EV vehicles are designed to handle high-frequency urban deliveries while maintaining low operational costs.
116 Inbound Logistics • January 2026
Powered by FlippingBook