Dear Snap4City user [current-user:og-membership--1]

TWO new positions for PHD course in DINFO University of Florence to work at DISIT lab for https://www.Snap4City.org area of interest on green tematic area

https://www.unifi.it/p12104

Deadline for applications 25th october 2021 at 12 (midday)

https://www.unifi.it/upload/sub/dottorati/PON_2021/All_2_schede_borse_ENG.pdf

see pages 35 and 36

-- Fleet management methods of e-vehicle, with machine learning techniques, explainable artificial intelligence and IoT, for the reduction of maintenance costs and environmental impact

We are observing a progressive growth of electric vehicles and their types, models and in particular of their use in fleets of rental vehicles or for city use for the use of operators, and therefore also of the related problems. These, having to manage significant numbers of vehicles, can control their evolution and maintenance, based on driving conditions, routes, and also the very structure of the mechanics and electronics of the vehicle. The primary objectives are the reduction of downtime for maintenance, and the reduction of unexpected failures that lead to emergency interventions, but also the management of refills, the identification of components that can fail, the profiling of periodic maintenance. These requests can be satisfied by developing Ai and XAI algorithms on the large amounts of data that are available. At the same time, the semantic modelling of the structures involved such as the vehicle itself, the maintenance processes, the roads travelled, the type of behaviour can guide and accelerate the learning processes. The study will exploit the infrastructure of https://www.Snap4City.org and the data of the DISIT lab unifi. 

-- What-if analysis methods for responding to unexpected environmental and non-environmental events, with explainable artificial intelligence and IoT techniques, to increase the resilience of urban and rural systems.

What-If analysis solutions have to cope with highly complex situations of city scenarios addressing unexpected events to increase resilience. The solutions have to be capable to compute multiple predictions and simulations about city evolution such as environmental variables, public transport, parking, people flow, commercial areas, etc. The approaches take into account data which are static, historical, real-time/dynamic, and forecasting information, in a functional model, on which the processes (simulations, predictions, data transformations) are integrated with business logic and user interaction. Despite the large literature of What-If analysis its complexity for managing actual cases of progressively computed results is far to be covered by solutions and tools. So that the classic prediction models cannot be used, since they have a limited performance to cope with unplanned events that have to be managed in a short time. Other relevant aspects to be addressed are the performance indicators to assess the results. The study is going to exploit the https://www.Snap4City.org infrastructure and data of the DISIT lab at Unifi. 

for drect contact please as to Paolo.nesi@unifi.it, also on skype, and paonesi@gmail.com