Key Limitations and Future Research

Limitations

  1. Lack of waste volume data: The research only had accumulated waste data for the entire district rather than individual collection points. This prevented the model from considering vehicle capacity constraints and load balancing.
  2. Absence of operational cost data: The optimization was based purely on distance minimization, without factoring in fuel costs, labour expenses, or vehicle maintenance, limiting the model’s real-world applicability.
  3. No data on actual depot routes: The existing waste collection routes used by the depot in Kuppiyawatha East were unavailable, making it impossible to compare the optimized path with real-world practices.
  4. Undirected graph assumption: The waste collection network was modeled as bidirectional, whereas real-world roads may have one-way restrictions, traffic conditions, and limited access zones that affect route optimization

Future Research

  1. Incorporating Waste Volume Data: Collecting real-time waste level data for each collection point would allow for dynamic vehicle load balancing, reducing unnecessary trips. This approach has been explored in studies utilizing ultrasonic sensors and IoT-based systems to monitor waste levels, enabling optimized collection routes (Long et al., 2019).
  2. Cost Optimization: Future models should consider fuel consumption, labor costs, and maintenance expenses, using multi-objective optimization techniques to balance cost and distance efficiency. Research has demonstrated the effectiveness of integrating cost factors into route optimization for waste collection (Kumar et al., 2022).
  3. Comparison with Real Depot Routes: By obtaining actual waste collection routes, future research can benchmark the optimized path against current operations to measure potential savings in time, distance, and cost. This comparison would provide practical insights into the benefits of the proposed optimization models.
  4. Directed Graph Modeling: Implementing one-way road restrictions and road conditions in the model would improve its realism and accuracy. Considering road network constraints is crucial for developing feasible and efficient waste collection routes (Muthuvel et al., 2024).
  5. Dynamic Routing Algorithms: Using real-time traffic and waste generation data, future models can implement adaptive route planning to respond to changing conditions. Studies have shown that dynamic route optimization can significantly enhance the efficiency of waste collection systems (Alharbi & Alshehri, 2020).
  6. Exploring Alternative Optimization Techniques: Algorithms such as A*, Genetic Algorithms (GA), and Ant Colony Optimization (ACO) could be tested and compared against Dijkstra’s Algorithm to find the best-performing approach for large-scale waste collection. Research indicates that these algorithms can offer advantages in solving complex routing problems (Karadimas et al., 2008).