We put great importance to data protection and therefore use the data you provide to us with upmost care. You can handle the data you provide to us in your personal dashboard. You will find our complete regulations on data protection and clarification of your rights in our privacy notice. By using the website and its offers and navigating further, you accept the regulations of our privacy notice and terms and conditions.
Shared mobility can contribute to greener and more livable cities. The limited knowledge about their adoption and use patterns is an obstacle for better service planning and management. Nommon is developing an AI-based tool that anticipates demand levels to optimise shared mobility operators.
Shared mobility services have recently emerged as a new transportation business model, enabled by digitalisation. These services offer shared on-demand vehicles that can be booked and used by citizens through mobile apps. Car sharing, moped sharing, bike sharing and e-scooter sharing services stand out as an alternative to private car use, often cleaner, as most of the fleets deployed are electric. In spite of their outstanding growth, these services still struggle to achieve profitability, hindering their financial sustainability. Hence, operators need to improve the efficiency of their fleet deployment and management strategies.
In order to overcome this challenge, a more comprehensive understanding of the adoption and use patterns of shared mobility services is required. Operators need to know more about the expected demand levels and the characteristics of the users. To meet this need, Nommon is developing a tool based on AI-powered models able to predict demand patterns that support operators in the planning and management of shared mobility services.
The tool leverages the large amount of data collected by shared mobility operators, which continuously monitor the position and use of their vehicles, thus acquiring an accurate description of the actual demand of the services. These datasets can be combined with other data sources which are able to describe the context in which this demand is captured. This is key to developing predictive models that are applicable beyond the current operation area, so that demand forecasts can be produced for supporting strategic decisions (e.g. implementation in a new city). In particular, the pervasive use of mobile devices has led to the availability of new data sources that register the activity and mobility patterns of the population (e.g. mobile network data, public transport smart card data). Fine-grained travel demand indicators can be extracted from these sources, such as origin-destination matrices at highly disaggregated temporal and spatial scales. The solution applied exploits all these datasets through a set of demand prediction machine learning algorithms. These models produce estimates of the captured demand for a given period of time at an origin-destination pair level, with a…
