Optimized Collaborative Routing for UAVs and Ground Vehicles in Integrated Logistics Systems
Abstract
1. Introduction
2. Related Study
3. Ant Colony Algorithm
Improved ACO with Two-Objective Optimization
Algorithm 1: Improved ACO to solve logistics vehicle travel path |
Input: Road network information matrix G, express time limit requirements, penalty coefficient. Output: Logistics vehicle travel path. 1: Initialize the parameters. The initial position of m ants is randomly distributed in each heavy component point. The first : To nullify, set the algebraic counter Nmax ← 300; pathway initial concentration P ← 1; information heuristic factor α ← 2; expectation heuristic factor β ← 5; the rate of evaporation of pheromone p to 0.1; and substitute the road network information matrix G into the distance table. 2: Randomly select the starting point s to join . 3: Set the current point of the logistics vehicle as w, and the next node set to be selected as . 4: For each point j in the next set of nodes to be selected, calculate the transfer probability, (t); select the next node j using roulette. . Update the paths added to the tab with local information until all heavy-duty points are added to a tab. 5: Judge whether the number of ants traversing the week has reached m, and if it has not reached it, perform step 2; otherwise, perform step 6. 6: Once the m-only ant colony solution is completed, seek the optimal path in the current round for global signaling update. The updated Equations are shown below: 7: Determine whether the cycle number n reaches 300, then implement step 8; otherwise, implement step 2. 8: Choose the global optimal solution according to the optimal solution of each round, export the logistics vehicle, and exit. |
4. Proposed Study
4.1. Road Network Construction
4.2. Road Network Construction Algorithm
Algorithm 2: Generate road network and road network information |
Input: Boundary limit parameters. Output: Intersection point set, road network information matrix. 1: Select the baseline, set the upper and lower boundaries and the left and right boundaries, and set the road level. 2: Randomly select several points with the right expansion set V = [1] and the edge set E = [1]. 3: Process the points in a cluster. 4: If the current point is n, the coordinates () are not processed; the nearest point above this point is (), the nearest subpoint is (), the upper critical point is (), and the lower boundary is (). According to the slope relationship between the upper and lower points and the processing points and the limit of the upper and lower boundary points, a new point is randomly generated: (), num + +. V = V ∪ point. 5: If this condition is not met, exit the current recursion. 6: Randomly generate a road congestion degree coefficient u, connect the new point and (), and generate a new side e: (n, num, path length ). distance[n][point] ←. E ← Eue. 7: If the new point () exceeds the boundary, exit the current recursion. 8: If not, add the new point to the point set, and return to step 3. |
4.3. Path Storage and Solving
Algorithm 3: Combine the Floyd algorithm to solve the storage of the shortest path length and shortest paths between any two points |
Input: Collection of intersection points V, direct road information E. Output: Shortest path length and shorter paths among any two points. 1: path[i][j] ← j, distance[i][j] ← . 2: For k ← 1 to Q Do. 3: For i ← 1 to Q Do. 4: For j ← 1 to Q Do. 5: if distance[i][j] > distance[i][k] + distance[k][j]. 6: distance[i][j] ← distance[i][k] + distance[k][j]. 7: path[i][j] ← path[i][k]. |
Algorithm 4: Solve the shortest path and shortest distance between two points |
Input: Node number a, b; Empty array A [1: ∞] Output: The shortest path length, the sum between intersection point a and intersection point b, the shortest path A [N]: 1: Sum ← 0; 2: N ← 0; 3: While a! = b Do; 4: Sum ← Sum + 1; 5: N ← N + 1; 6: A[N] ← a; 7: a ← path[a][b]; 8: Output Sum, and A [N]. |
4.4. Regional Segmentation Labeling
5. Drone–Logistics Vehicle Co-Planning Algorithm
5.1. Collaborative Algorithm Analysis
5.2. A Drone Flight Path Planning Algorithm
Algorithm 5: Calculate the flight path of the UAV on the light point |
Input: GV driving order, the number of UAV, light point collection light = ; Route Network Information Matrix G (). Output: Flight paths of all light points in the UAV pair area. 1: Initially, according to the GV driving order, call Algorithm 2 and Algorithm 3 to get the GV-specific driving path, A [N] = . 2: Light traversal, for each element in the light for the nearest direct path from e = . The distance is d. Set the nearest point on the road e to as the landing point of the UAV end. 3: Calculate the take-off point and landing point of the UAV for each light point. Determine the departure point . Start satisfies the following equation: = (straight line distance to start) fls = (start of the road distance) 4: Sort according to the take-off and landing points of the lightweight points. 5: If the combined weight of the adjacent lightweight points is less than the maximum load of the drone, the flight routes of the two are combined. Merge strategy is as follows: 6: for i ← 1 to | light | −1. |
- The customer waiting time cost, based on the time, elapsed between dispatch and successful delivery at each point.
- The driving cost incurred by the logistics vehicle throughout the delivery route.
- The flight cost associated with the drone’s operations, considering both distance and number of sorties.
6. Results and Discussion
- The coordinates of each delivery point.
- The corresponding delivery time window.
- The penalty coefficient for late delivery.
- Greedy Algorithm;
- Traditional ACO;
- Improved ACO (proposed in this study).
6.1. Optimization Quality Comparison
6.2. Comparison with Traditional Distribution
6.3. Comparison Between Traditional and Drone-Assisted Delivery
7. Conclusions
- A 16.5% average driving cost reduction;
- An 8.3% average waiting time reduction;
- A 22.7% energy savings for UAV operations;
- A 14.9% reduction in total delivery time.
Author Contributions
Funding
Data Availability Statement
Conflicts of Interest
Abbreviations
P | Number of UAVs |
K | Maximum payload of UAV |
V | Set of all nodes |
Vs | Set of points the logistics vehicle passes through |
Vt | Collection of all express points |
X | Real-time location with dynamic changes |
E | Set of all direct paths |
T | Real time |
MILP | Mixed-Integer Linear Programming |
TSP | Traveling Salesman Problem |
GV | Ground vehicle |
Ti | Time the; ogistics vehicle visits point i |
Time the drone visits point i | |
Latest arrival time at the express point i | |
Driving time between nodes i and j | |
Number of flights with UAV i | |
Total weight of the express points i | |
Service time of the express point j | |
Subset of edges traversed by GV | |
Subset of edges traversed by the UAV | |
GV unit distance travel cost | |
UAV unit distance travel cost | |
GV driving speed | |
UAV flight speed | |
Distance of the ith flight of the jth UAV | |
Penalty coefficient of express point i | |
Road distance between the ith and jth points | |
m | Number of nodes in the network is |V| + 1 |
n | Number of express points, the value is || |
drf | Driving cost impact coefficient |
waf | Customer waiting time impact coefficient |
GV is 1 if it travels from point i to point j | |
UAV takes off from the ith point, jth point |
References
- Agatz, N.; Erera, A.; Savelsbergh, M.; Wang, X. Optimization for dynamic ride-sharing: A review. Eur. J. Oper. Res. 2012, 223, 295–303. [Google Scholar] [CrossRef]
- Aras, N.; Aksen, D.; Tekin, M.T. Selective multi-depot vehicle routing problem with pricing. Transp. Res. Part C Emerg. Technol. 2011, 19, 866–884. [Google Scholar] [CrossRef]
- Bektaş, T.; Laporte, G. The pollution-routing problem. Transp. Res. Part B Methodol. 2011, 45, 1232–1250. [Google Scholar] [CrossRef]
- Bakirci, M. A Novel Swarm Unmanned Aerial Vehicle System: Incorporating Autonomous Flight, Real-Time Object Detection, and Coordinated Intelligence for Enhanced Performance. Trait. Du Signal 2023, 40, 2063–2078. [Google Scholar] [CrossRef]
- Asaamoning, G.; Mendes, P.; Rosário, D.; Cerqueira, E. Drone swarms as networked control systems by integration of networking and computing. Sensors 2021, 21, 2642. [Google Scholar] [CrossRef] [PubMed]
- Rismanto, H.; Judijanto, L. Dynamic Routing in Urban Logistics: A Comprehensive Review of AI, Real-Time Data, and Sustainability Impacts. Sinergi Int. J. Logist. 2025, 3, 68–79. [Google Scholar] [CrossRef]
- Oliveira, C.A.; Pardalos, M.P. Handbook of Artificial Intelligence and Data Sciences for Routing Problems; Springer: Berlin/Heidelberg, Germany, 2025. [Google Scholar]
- Holland, J.H. Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence; MIT Press: Cambridge, MA, USA, 1992. [Google Scholar]
- Cachon, G.; Terwiesch, C. Matching Supply with Demand: An Introduction to Operations Management; McGraw-Hill: Singapore, 2009. [Google Scholar]
- Erdoğan, S.; Miller-Hooks, E. A Green vehicle routing problem. Transp. Res. Part E Logist. Transp. Rev. 2012, 48, 100–114. [Google Scholar] [CrossRef]
- Cordeau, J.-F.; Laporte, G.; Mercier, A. A unified tabu search heuristic for vehicle routing problems with time windows. J. Oper. Res. Soc. 2001, 52, 928–936. [Google Scholar] [CrossRef]
- Liu, X.-F.; Dai, Y.-T.; Fang, Y.; Yu, X.; Zhan, Z.-H.; Zhang, J. Fragment-Based Knowledge Transfer for Multi-Task Capacitated Vehicle Routing. IEEE Trans. Intell. Transp. Syst. 2025, 1–15. [Google Scholar] [CrossRef]
- Xue, Z.; Xu, H.; Cui, H.; Peng, W. A machine learning-incorporated heuristic column generation algorithm for truck grouping and scheduling problem in trailer-swapping transport mode. Expert Syst. Appl. 2025, 278, 127327. [Google Scholar] [CrossRef]
- Rehman, S.; Hassan, H. Empirical Assessment of National Logistics Performance and Its Impact on Advancing the Sustainable Development Goals: A Global Analysis. Available online: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5200258 (accessed on 20 November 2024).
- Hammad, F.; Sattar, F.A.; Ahmed, W.; Akram, M.A. Understanding the influence of strategic reverse logistics management on the sustainable performance of the organization. OPSEARCH 2025, 1–21. [Google Scholar] [CrossRef]
- Aranguiz, A.A.; Spoelstra, M. Trends and Outlook of Dairy Production in Pakistan; Wageningen Livestock Research: Wageningen, The Netherlands, 2025. [Google Scholar] [CrossRef]
- Ahmad, M.; Waqar, Z. Towards Sustainable Health Systems: Insights From Pakistan’s Spatial Data Infrastructure. In Exploring Pillars of Sustainability for Modern Age Improvements; IGI Global Scientific Publishing: Hershey, PA, USA, 2025; pp. 373–390. [Google Scholar]
- Aralica, Z.; Škrinjarić, B.; Roud, V. How sectoral technological transformation explains firm-level productivity? Post-Communist Econ. 2025, 1–22. [Google Scholar] [CrossRef]
- Xia, Y.; Yin, H.; Mohsin, M.; Mehak, A.; Yan, C. Exploring FTAs, seafood exports, and SDGs: A gravity model analysis of Pakistan’s seafood trade with China and regional partners. Front. Mar. Sci. 2025, 12, 1553780. [Google Scholar] [CrossRef]
- Jamil, K.; Zhang, W.; Anwar, A.; Mustafa, S. Exploring the Influence of AI Adoption and Technological Readiness on Sustainable Performance in Pakistani Export Sector Manufacturing Small and Medium-Sized Enterprises. Sustainability 2025, 17, 3599. [Google Scholar] [CrossRef]
- Anwar, M.W.; Mustafa, G.; Taimoor, M.; Akram, M.R. An Analysis of Pakistan-Saudi Arabia Relations in the 21st Century. Compet. Res. J. Arch. 2025, 3, 232–242. [Google Scholar]
- Zhou, X.; Zhao, C.; Xie, F. Collaborative Delivery Route Planning for Rural Public Buses and Express Delivery Vehicles Based on Passenger and Freight Integration. Transp. Res. Rec. 2025, 03611981251318333. [Google Scholar] [CrossRef]
- Korani, E. Estimation of Route Reliability in Multimodal Hierarchical Hub Location Problem Using Lagrangian Relaxation and Artificial Neural Networks (ANN). Res. Ann. Ind. Syst. Eng. 2025, 2, 48–71. [Google Scholar]
Ant Colonies Forage for Food | Ant Optimization Algorithm | Applications | Key Challenges |
---|---|---|---|
Ant colony | A set of efficient solutions in the search space (expressed as population size N) | Traffic route, logistics optimization | Balancing exploration vs. exploitation |
Foraging space | Search space of the question represented by dimension D | Warehouse management, network design | High-dimensional complexity |
Pheromone | Information concentration variable | Robotics colony intelligence | Decay and dynamic updates |
A path from the nest to the food | An effective solution | Supply chain efficiency | Local optima traps |
Shortest path found | Optimal solution of the problem | Route optimum | Scalability in large network |
Point Serial Number | Coordinate x | Coordinate y | Demand Time (S) | Penalty Coefficient | Location x | Request Time (S) |
---|---|---|---|---|---|---|
0 | 687 | 252 | 0 | 0 | 687 | 0 |
1 | 890 | 304 | 6 | 4 | 890 | 6 |
2 | 569 | 717 | 4 | 3 | 569 | 4 |
3 | 858 | 353 | 15 | 2 | 858 | 15 |
4 | 11 | 118 | 28 | 1 | 11 | 28 |
5 | 53 | 285 | 26 | 3 | 53 | 26 |
6 | 746 | 196 | 4 | 4 | 746 | 4 |
7 | 204 | 19 | 20 | 4 | 204 | 20 |
8 | 526 | 125 | 10 | 1 | 526 | 10 |
9 | 522 | 268 | 11 | 3 | 522 | 11 |
Driving System Number | Time (S) | Improved ant Colony | Ant Colony | Greedy Algorithm | Solution Quality (Ratio) | Computitional Efficiency (Ratio) |
---|---|---|---|---|---|---|
0.1 | 0.9 | 5139.27 | 6395.33 | 9553.38 | 0.330569 | 0.244404 |
0.2 | 0.8 | 4719.36 | 5818.58 | 8857.62 | 0.455996 | 0.232917 |
0.3 | 0.7 | 4427.08 | 5432.88 | 8161.86 | 0.448915 | 0.227193 |
0.4 | 0.6 | 4177.19 | 5014.56 | 7466.1 | 0.328356 | 0.200463 |
0.5 | 0.5 | 3596.87 | 4235.49 | 6770.34 | 0.430385 | 0.177549 |
0.6 | 0.4 | 3335.25 | 3835.79 | 6074.58 | 0.417937 | 0.150079 |
0.7 | 0.3 | 3215.1 | 3633.79 | 5378.82 | 0.324426 | 0.130226 |
0.8 | 0.2 | 2894.4 | 3173.54 | 4683.06 | 0.322338 | 0.096438 |
0.9 | 0.1 | 2573.71 | 2713.27 | 3987.3 | 0.319522 | 0.054225 |
Group | Light Item Ratio (%) | Time (S) | Traditional Delivery Driving Cost (USD) | Drone Time (S) | Assisted Driving Cost (USD) | Volume Light Item | Volume Heavy Item |
---|---|---|---|---|---|---|---|
1 | 14.28 | 147,383.51 | 2664.29 | 1,036,493.139 | 2477.32 | 66 | 11 |
2 | 26.09 | 21,499.08 | 3121.79 | 1,424,407.768 | 3076.11 | 85 | 30 |
3 | 21.38 | 323,904.68 | 3663.48 | 2,249,596.078 | 3333.97 | 119 | 33 |
4 | 21.81 | 544,571.67 | 4314.34 | 3,579,578.157 | 4099.99 | 147 | 41 |
5 | 20.74 | 684,587.45 | 4675.86 | 3,846,336.020 | 4316.55 | 172 | 45 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Rashid Nazir, H.M.; Sun, Y.; Hu, Y. Optimized Collaborative Routing for UAVs and Ground Vehicles in Integrated Logistics Systems. Drones 2025, 9, 538. https://doi.org/10.3390/drones9080538
Rashid Nazir HM, Sun Y, Hu Y. Optimized Collaborative Routing for UAVs and Ground Vehicles in Integrated Logistics Systems. Drones. 2025; 9(8):538. https://doi.org/10.3390/drones9080538
Chicago/Turabian StyleRashid Nazir, Hafiz Muhammad, Yanming Sun, and Yongjun Hu. 2025. "Optimized Collaborative Routing for UAVs and Ground Vehicles in Integrated Logistics Systems" Drones 9, no. 8: 538. https://doi.org/10.3390/drones9080538
APA StyleRashid Nazir, H. M., Sun, Y., & Hu, Y. (2025). Optimized Collaborative Routing for UAVs and Ground Vehicles in Integrated Logistics Systems. Drones, 9(8), 538. https://doi.org/10.3390/drones9080538