LoRaWAN Ultrasonic Smart Waste Bin Monitoring in Mumbai

Fixed-route collection assumes every bin fills at the same rate. They don’t. In practice, the same route was too slow for the busiest bins and too frequent for the quietest ones.

The cost of that mismatch was paid twice. Overflowing bins meant litter on roads, stray animals, foul smell and citizen complaints, which are the most visible failures a city can have. Trips to half-empty bins meant fuel, driver hours and vehicle wear spent on work that didn’t need doing.

Underneath both problems sat one gap: nobody knew a bin’s status until someone physically looked at it.

Each bin was fitted with a MacRay-LU4 LoRaWAN ultrasonic level sensor, mounted under the lid and pointing down at the waste. Here’s how it works:The sensor sends an ultrasonic pulse and times the echo from the waste surface to measure distance.

1. It converts that distance into a fill percentage using the bin’s own depth.
2. It sends the reading over LoRaWAN, a long-range, low-power wireless network, to gateways covering the service area.
3. The dashboard plots every bin on a map and alerts supervisors the moment a bin crosses its threshold.

Why LoRaWAN rather than GSM/4G sensors? A bin only needs to send a few bytes a few times a day. LoRaWAN does that on India’s licence-free IN865 band with no SIM card or data plan per bin, and it uses far less power than a cellular modem. That is what makes multi-year battery life, and therefore city-wide scale, practical.

How reliable the readings are depends on three installation decisions.

1. Mount at the centre of the lid, pointing straight down.
An off-centre sensor sees the bin wall before it sees the waste, which gives falsely high readings.

2. Calibrate every bin type, not just one.
Fill level is calculated as:

Fill % = (Empty distance − Measured distance) ÷ (Empty distance − Full distance) × 100

• Empty distance is from the sensor to the bin floor.
• Full distance is the point you define as full. Set it a few centimetres below the lid, clear of the sensor’s 3 cm blind zone.

Each bin type was set up with its own values using the Macnman Maya app over Bluetooth.

3. Set the alert threshold by response time, not habit.
A bin should alert early enough for a crew to reach it before it overflows:

Alert threshold ≈ 100% − (fill rate per hour × crew response time in hours)

Example: a market bin that fills about 10% per hour, with a crew that needs 3 hours to arrive, should alert at about 70%. A residential bin filling 2% per hour can safely alert at 90%.

Why ultrasonic here, and ToF elsewhere? In our warehouse waste-room project we used an 8×8 Time-of-Flight sensor, because large open rooms fill very unevenly from corner to corner. A closed street bin is the opposite case: a small, fixed-size container exposed to monsoon rain, heat and dust. There, a rugged ultrasonic sensor with a wide 60° beam gives dependable readings at low cost per bin. Weight-based sensors were not used, because light, bulky waste such as packaging and thermocol can fill a bin while barely changing its weight.