Project summary

• Unplanned machine downtimes in industry cause significant production losses and costs. This project proposes the design and implementation of an Intelligent IoT system for continuous monitoring, data collection and intelligent analysis to anticipate or quickly detect machine stoppages.
• Reference: SEE26-IOT-03. The solution will combine sensors, edge processing, connectivity, cloud storage/processing and alerting mechanisms for industrial machines.

Main objectives

• Design and prototype an IoT-based monitoring solution capable of detecting and alerting on machine downtime and anomalous behavior in real time.
• Implement intelligent analysis (rule-based and/or ML-based) to anticipate failures or rapidly detect stoppages and trigger alerts to operators.

Technical tasks and activities

• Sensor and hardware selection: evaluate and select sensors (vibration, current, temperature, acoustic, encoder, etc.), microcontroller or edge-compute platform, and power/communication options.
• Edge software: develop firmware for data acquisition, preprocessing, local anomaly detection or feature extraction, and reliable communication (MQTT/HTTP/CoAP) to backend/cloud.
• Backend and analytics: design data pipeline, storage, dashboards and implement analytics/ML models for downtime detection and predictive alerts.
• Alerting and integration: implement alert channels (SMS, email, push, dashboard notifications) and mechanisms to integrate with existing industrial systems (SCADA/OPC-UA where applicable).

Candidate profile and required skills

• Background in IoT / Embedded Systems, Industrial Automation, or Computer Engineering.
• Practical skills in: microcontrollers or SBCs (e.g., ESP32, Raspberry Pi), sensor interfacing, Python/C/C++, MQTT and RESTful APIs, data processing and ML basics for anomaly detection.
• Familiarity with cloud platforms (AWS/Azure/GCP) or time-series databases and dashboarding tools (Grafana, InfluxDB) is a plus.

Expected deliverables

• Working prototype (hardware + firmware) able to monitor at least one type of industrial machine and detect downtime events.
• Backend components: data ingestion pipeline, a basic dashboard for visualization, and alerting mechanism.
• Documentation: installation and user guide, technical report describing design choices, dataset collected during tests and evaluation results of detection accuracy.

How to apply

• To apply for this PFE, use the online application link: Apply here
• In your application include a brief CV, a motivation letter specifying relevant skills/experiences, and any prior project work related to IoT or predictive maintenance.