One of the main difficulties faced by shared vehicle companies and public transport is to maintain the vehicle interior in optimal conditions when the user changes, to improve the experience of the journey.
As part of the AntiTrash project, a system will be developed to automatically detect valuables, rubbish or potential damage in the shared vehicle or public transport. The system will produce an estimate of the cleanliness so that the transport operator can take necessary action. There are cameras inspecting the interior of the vehicle or public transportation, and through machine learning techniques, images are used to assess its state. This will also enable identification of the person responsible for the damage and/or the rubbish within the shared vehicle so that they can be directly charged maintenance costs, if necessary.
In addition, smart materials that increase the durability of the interior are promoted (plastics, metals and polymers), as well as the use of nanotechnology (microfluidic hydraulic systems) for detecting smells and particles, to ensure the good quality of air inside the car. This also contributes to reducing cleaning costs and makes it easier to replace vehicle components.
The AntiTrash project is funded by EIT Urban Mobility and is expected to continue until the end of 2021. The project is coordinated by Aalto University, and other participants include Carnet, the city of Hamburg, Electrobus Europe, Hamburger Hochbach, Municipality of City of Zalaegerszeg, Technische Universität Braunschweig - Niedersächsisches Forschungszentrum Fahrzeugtechnik (NFF), Plezintor, Seat, Technical University of Catalonia, Tusgsal, and Zone Cluster.
AntiTrash tested in the German subway and Hungarian busses
The AntiTrash camera units developed at Aalto University are currently being tested in collaboration with Hamburger Hochbahn and Ikarus busses in Hungary. In Hamburg a subway train on the U2-line has been equipped with cameras, collecting image data each time the train reaches one of the end stops of the line. Capturing data from the real application environment is crucial for properly designing the system.
The captured image data is utilized to validate the camera-based approach for interior condition monitoring, as well the detection accuracy of the applied algorithms. Furthermore, the algorithms are improved overtime with the help of the statistics acquired from the image data. Future developments will look into more precisely recognizing the state of cleanliness in the cart and busses, and offering convenient solutions for service providers to monitor the condition of their vehicles.
AntiTrash development environment
In the short video below you can see the AntiTrash development environment at the Aalto Industrial Internet Campus.