Integration of the sustAGE system

The sustAGE platform brings together an IoT ecosystem which is deployed in industrial environments, specifically in manufacturing (indoors) and ports (outdoors) and a set of software components that aim to increase productivity and wellness of ageing employees by monitoring their activities and state considering spatio-temporal aspects. Beyond the work environment the system will support daily activities of the user through user-context monitoring information from a subset of sensor data. To integrate such a diverse set of technologies and components, an agile way of working has been setup. Based on agile principles, we have setup frequent integration cycles, rapid prototyping and close collaboration between self-organizing, cross-functional teams. Moreover, a privacy-by-design approach has been followed. No personally identifiable information is sent to the analysis components so as to preserve privacy and ensure that identities of the participating workers will not be disclosed to unauthorized parties, including system operators and developers. Furthermore, the first stage processing of raw data collected by the sensors is performed on-premises and close to the original source in order to preserve privacy restrictions but also to improve performance of the analysis since the data stream doesn’t have to endure network latencies before being processed.

The following schema depicts the 4 layers of the architecture: Monitoring Layer, Streaming Layer, Personalisation Layer and Recommendation Layer and their interconnections.

The sustAGE Architecture

Upon collection and first-stage analysis of the sensor data in the Monitoring Layer, a message-based communication with the rest of the system components takes turn. The Streaming Layer contains the services to allow components to subscribe or publish messages to the so-called topics which represent channels of information that is circulated in real-time. The Personalisation Layer processes the incoming messages, detects key user micro-moments and produces the necessary user-specific knowledge that will help the system to build a personalized recommendation framework. The latter is supported by the Recommendation Layer components that deliver the produced recommendations to workers. This delivery takes place via different channels, i.e. via mobile notifications and wristwatch alerts. User response to the received recommendations is sent back to the system which constantly adapts to every user’s feedback so as to improve the personalized recommendations.

We are currently developing the first integrated version of the sustAGE platform, the so-called Minimum Viable Product (MVP). The MVP will demonstrate the above capabilities in a few baseline scenarios of usage and will also help the consortium receive feedback from users early enough so as to undertake any required adjustments in view of the release of the 1st integrated prototype.

We are currently developing the first integrated version of the sustAGE platform, the so-called Minimum Viable Product (MVP). The MVP will demonstrate the above capabilities in a few baseline scenarios of usage and will also help the consortium receive feedback from users early enough so as to undertake any required adjustments in view of the release of the 1st integrated prototype.