AIPEX | AI-assisted IoT devices production and exploitation

Scaling smart IoT for predictive maintenance

Predictive maintenance relies on IoT sensors, but large-scale deployment remains limited. AIPEX supports the I-Care company in transitioning to industrial-scale production, with a target to increase volumes by a factor of 100.

The project addresses two key challenges: the availability of critical components such as accelerometers and the efficient management of large sensor fleets. It develops solutions to industrialise production and optimise sensor operation.
 

Target group

The project primarily targets I-Care and Sagacify, partners in the project.

It will also benefit industrial clients (manufacturing, chemicals, pharma, food & beverage) whose critical infrastruc-ture and equipment can benefit from predictive maintenance solutions.

Context

Today, predictive maintenance and IoT production face several limitations. High-performance sensors remain dif-ficult to produce at scale, while production lines are still too manual and costly.

At the same time, managing and analysing data from large sensor fleets is complex. Existing solutions lack integra-tion with AI tools to optimise both production and operations.

These limitations hinder the development of scalable industrial models such as “Product-as-a-Service” and call for new approaches combining hardware innovation and advanced AI.
 

Objectives & results

AIPEX aims to remove technological barriers related to the large-scale production and operation of smart IoT sensors.

The benefits that the partner companies (I-Care and Sagacify) will enjoy are:

• A high-performance accelerometer compatible with mass production
• An IoT predictive maintenance solution optimised for industrialisation and cost efficiency
• An AI platform to optimise production parameters (cost, quality, throughput)
• A platform to manage millions of sensors using distributed intelligence
• A scalable and industrialised manufacturing process

Approach

The project is structured into several complementary workstreams combining hardware development, artificial intelligence and industrial validation:

1. Hardware development

• Design and validation of a next-generation accelerometer
• Redesign of the IoT product for mass production

2. Production optimisation (Manufacturing Optimization Platform)

• Collection and integration of production data
• Development of reinforcement learning algorithms
• Simulation and optimisation of production lines

3. Lifecycle management (Sensors Management & Tracking Platform)

• Development of federated learning models
• Analysis of sensor data
• Anomaly detection and predictive maintenance

4. Industrialisation & validation

• Testing on real production lines
• Integration of solutions in an industrial environment
   

Funding

• Funding agency: SPW
• Project type: Pôle Mécatech