Developing a Smart Wireless Sensor Network for Waste Management in Cities

This paper proposes and evaluates a city-scale wireless sensor network (WSN) for smart waste management that combines IPv6-enabled IoT nodes, the RPL routing protocol, and a UAV-based collection strategy. We simulate a realistic urban scenario based on transport stations around Whitechapel, London, and publish the full Contiki-NG/Cooja simulation and node source code for reproducibility.

Key ideas and architecture:

• Three node roles:

  • Root node (onboard a UAV): collects bin status, prioritizes the fullest bin, and schedules emptying in 30-minute cycles.

  • Waste nodes: smart bins that sense fill level and periodically send status.

  • Relay nodes: placed strategically to maintain multi-hop connectivity in dense urban settings.

• Networking stack:

  • RPL over IEEE 802.15.4 using 6LoWPAN, enabling IPv6 and multi-hop reliability in low-power, lossy environments.

• Case study:

  • Whitechapel transport stations serve as proxy locations for deploying 22 nodes in simulation.

  • Cooja (Contiki-NG v4.9) used with a controlled channel model and TMote Sky motes.

Experiments and findings:

• Transmission delays (22-node topology, 10-hour sim):

  • Average end-to-end delay ≈125 ms, with most nodes around 100 ms.

  • Delays increase with Euclidean distance from the root due to additional hops.

  • Occasional drops (e.g., node 18) highlight the need for retransmission/fault tolerance, but impact is minor given non-time-critical scheduling.

• Waste management behavior (5-node scenario, 10-hour sim):

  • A simple linear growth model with random rates (1–5%) per interval.

  • UAV/root selects the fullest bin at each 30-minute interval and “empties” it gradually, demonstrating effective prioritization.

Why it matters:

• Integrates proven IoT networking (RPL/6LoWPAN) with UAV logistics to reduce emissions and improve service efficiency.

• Demonstrates tolerable latency for periodic monitoring in urban deployments.

• Offers open-source code and simulations to accelerate future research and real-world pilots.

Limitations and future work:

• The waste growth model is simplified; real environments need richer, non-linear models and human/temporal patterns.

• Packet loss handling and reliability mechanisms (e.g., retransmissions, acknowledgements, fault tolerance) should be added.

• Physical pilots and AI-based scheduling/route optimization could further enhance performance and adaptivity.

Resources:

• Full simulation environment and code: https://github.com/palzino/Waste-Management-WSN
• DOI: https://doi.org/10.1007/978-3-031-91351-8_12