top of page
20180409 DataCity new logo2-09.png

- PROJECTS -

Smart escalators

How is it possible to reduce the downtime of a station escalator to improve customer satisfaction?

THE CONTEXT


Stations are the access point to the railway system. Before boarding a train, travellers almost always use the escalators, which are critical for good passenger flows. Availability is crucial for a smooth journey and to avoid crowds building up on platforms. The aim is to ensure that enough escalators are available for unimpeded passenger flow in stations, in order to meet the objectives set by transport organisation authorities such as Île-de-France Mobilités.

 

99%

target set by SNCF for availability

of escalators in Île-de-France

 

THE CHALLENGE


SNCF Gares & Connexions and Bouygues Energies et Services have chosen to work with Fieldbox.ai, a startup specialised in implementing artificial intelligence in industry, in order to design a predictive maintenance system for escalators. By installing sensors on the equipment, they are able to monitor performance parameters in real-time and then correlate them with actual breakdowns. The solution which they developed can therefore alert technicians and maintenance staff of possible breakdowns.

 

THE STARTUP


The FieldBox.ai solution, which is the fruit of several years of R&D, is a cloud-to-edge IoT platform, which manages the lifecycle of these industrial bots: training, rollout and autonomous operation. By relocating decision-making to factories and automating it with AI, FieldBox.ai drastically cuts the cost and time required to achieve Industry 4.0.



 

THE PARTNERS


 

THE SOLUTION


Within the framework of the experimentation, a data capture tool, which can be accessed remotely, was delivered. A cloud version of the FieldBox.ai software was rolled out and configured for the requirements of the project. An operations dashboard displays the results. Initial analytics for the relationships between the various criteria makes it possible to suggest an approach to implementing a predictive model for breakdowns and recommendations for action.

 


Comments


bottom of page