IoT Hostel Washing Machine Usage Monitor

Shared washing machines in hostels and apartment complexes are a classic commons problem: a handful of machines serve hundreds of residents, there is no booking system, and the only way to know if one is free is to walk to the laundry room. Machines run overloaded, nobody knows which one is due for service, and "someone left their clothes in for three hours" is a daily complaint. This project instruments each machine with a compact sensor node that recognizes the wash cycle from its vibration and current signature — fill, agitate, spin phases are mechanically distinct — and reports state changes over Wi-Fi. The dashboard presents a live availability board (FREE / WASHING / CYCLE DONE), per-machine cycle counts and run-hours, a done-alert feed, and flags for abnormal behavior like a machine vibrating far beyond its normal band or a cycle running far too long. The student delivers the sensing hardware and a genuinely working dashboard demo that mirrors the real deployment.

1. A sensor node is mounted on each washing machine's body (vibration sensor) with a CT clamp on its supply cable — no panel is opened.
2. Firmware continuously watches vibration energy and current draw, classifying the machine into IDLE, FILL, WASH, SPIN or DONE with a phase state machine and debounce timing.
3. State changes are published immediately over Wi-Fi (MQTT/HTTP); a periodic heartbeat confirms each node is alive.
4. The backend accumulates cycle counts, run-hours and phase durations per machine; abnormal vibration or over-long cycles raise maintenance flags.
5. The dashboard renders the live availability board, per-machine usage stats, the done-alert feed and the maintenance flag list.
6. The student commissions the laundry room with the included procedure: run one supervised cycle per machine to learn its signature, set the phase thresholds, and validate the board against observed cycles.

Tech stack:

• ESP32 sensor nodes (one per machine)
• Vibration sensor + SCT-013 CT clamp per node
• Wi-Fi telemetry (MQTT/HTTP)
• Web dashboard (HTML/CSS/JS demo included)
• Phase state-machine firmware
• Arduino IDE (C/C++ firmware)
• 5 V USB power per node

• [Live availability board] Each machine shows FREE, WASHING or CYCLE DONE in real time — residents check their phones instead of walking to the laundry room.
• [Cycle-phase state machine] Firmware recognizes fill, wash, rinse and spin phases from the vibration/current signature, so the board reflects reality, not a timer guess.
• [Cycle-done alerts] The moment a machine finishes, the dashboard flags it — cutting the "forgotten clothes" idle time that blocks the next user.
• [Per-machine usage logs] Cycle counts and run-hours per machine give wardens the data behind maintenance scheduling and replacement decisions.
• [Abnormal-behavior flags] Excessive vibration or abnormally long cycles raise maintenance alerts with timestamps — the early warning a warden never had.
• [Non-invasive sensing] Nodes mount externally and sense vibration plus supply current via CT clamp — no machine is opened and no warranty is voided.
• [Working demo dashboard] A single-file web dashboard with realistic simulated laundry telemetry is included for presentation and viva.

• Sensor node design (one node per machine in the demo set)
• Phase state-machine firmware with debounce logic
• Working single-file dashboard demo with simulated laundry telemetry
• Mounting and signature-learning commissioning guide
• Project report PDF (theory, sensing, state machine, dashboard, results)
• PPT presentation and viva Q&A document

• Phase recognition is signature-based and approximate; heavily unbalanced loads or a machine with a failing damper produce unusual signatures the state machine may misread — stated honestly.
• Thresholds must be learned per machine model; one laundry room with mixed models needs per-machine commissioning.
• The dashboard demo ships with simulated telemetry; live data appears only after nodes are mounted and connected.
• Wi-Fi must cover the laundry room; the node buffers briefly but is not a long-term offline logger.
• It monitors machine state, not laundry ownership — it cannot tell whose clothes are inside.
• Cycle-done alerts appear on the dashboard; push notifications to phones need the notification extension listed as future scope.