Robot Assisted Truck for Food and Medication Delivery
Traditional last-mile delivery systems face significant inefficiencies, particularly in some communities where access to food and medication is limited by logistical and infrastructural constraints. The “last mile”, or more specifically, the “last 50 feet”, is often the most costly and challenging segment of the supply chain due to factors such as traffic congestion, labor shortages, difficult terrain, and lack of delivery infrastructure. In rural areas, these challenges are magnified by low population density, longer distances, and inadequate road networks. In urban neighborhoods, narrow streets and limited truck access create additional barriers.
Recent advancements in autonomous ground delivery robots and drones have opened new avenues for rethinking last-mile logistics. Cities like Miami and Detroit have begun testing robot delivery systems to reduce delivery time, improve reliability, and lower operational costs. These systems demonstrate the potential of automation in mitigating common urban delivery challenges. However, current implementations primarily focus on dense urban environments with robust technological and infrastructural support, leaving out the unique challenges present in less-connected or rural settings. Robot-Assisted Truck Delivery (RATD) systems introduce a hybrid logistics model where traditional vehicles work in tandem with autonomous robots. This collaboration offers multiple advantages: reduced delivery times, minimized traffic-related delays, and lower labor and fuel costs. Furthermore, robot deployment from moving vehicles expands the serviceable area and improves scalability without requiring full autonomy over long distances. These innovations are particularly critical in enhancing the delivery of food and medication. The proposed framework has the potential to serve as a replicable model for local agencies, transit authorities, and nonprofits seeking to deploy cost-effective last-mile solutions.
In this proposal, traditional vehicles and robots work collaboratively, with robots/drones dynamically joining or leaving groups to optimize deliveries. Robots, with their limited size and weight capacities, can ride on traditional vehicles for efficiency, launching from vehicles to make deliveries and then returning for recharging. Recent advancements in robotics have laid the technical groundwork for such collaborations, but there’s a lack of research in transportation and operations research communities on decision-making processes for vehicle-and-robot collaboration. This gap includes holistic optimization methods for planning and scheduling robotic fleets and coordinating robot launch and retrieval with moving vehicles.
To address these gaps, the research team proposes the following solutions:
1. Developing continuous approximation formulations and solution algorithm to optimize vehicle-and-robot delivery operations, with consideration of vehicle and robot synchronizing operations.
2. Designing deep learning models to predict key performance metrics, such as delivery costs, robot capacity, battery.
3. Developing decision-making frameworks to facilitate seamless collaboration between vehicles and robots, ensuring synchronized operations and optimized coordination for maximum efficiency.
Overall, this project aims to bridge the gap in achieving synchronized operations and coordinated deployment between traditional vehicles and robots. By optimizing their collaboration, the research seeks to improve food availability and overcome key logistical challenges.
The research team from Florida State University and Arizona State University will partner with Second Harvest of the Big Bend to improve food delivery operations through a Robot-Assisted Truck Delivery System. This system integrates traditional vehicles and autonomous robots to enhance food accessibility. Through close collaboration, the team and Second Harvest will share and refine operational strategies to ensure the system is effectively tailored to meet specific delivery needs and operational goals.
Second Harvest of Big Bend will provide valuable input on the logistical challenges and food distribution needs, helping the research team tailor the system's design to real-world conditions. This partnership will also play a key role in testing and validating decision-making frameworks that ensure the system is effective and integrates seamlessly into existing infrastructure. The community partner’s involvement ensures that the solutions proposed are both practical and responsive to the operational realities faced by food distribution organizations.
To optimize food delivery, the research team will develop and refine operational strategies for routing, scheduling, and coordinating both vehicles and robots. The focus is on reducing delivery times, improving efficiency, and addressing the logistical challenges. Decision-making frameworks will be designed to enable seamless collaboration between vehicles and robots, whether operating sequentially or concurrently. These frameworks will incorporate machine learning algorithms to predict key performance indicators such as delivery cost, robot capacity, and scheduling constraints, allowing the system to adapt dynamically and operate at peak efficiency. Ongoing feedback sessions and pilot implementations with Second Harvest will ensure the solutions remain practical, scalable, and responsive to real-world needs.
TBD