Menu Close
Close

HOW DOES MITHOS PLATFORM WORK?

1. DATA GATHERING

First, MITHOS platform receives a large amount of heterogeneous data from different internal and external sources (i.e., historical databases, infrastructure sensors and users) as well as from the scenario definition. This data would be provided from several information system (even from IoT devices) and with different format types (such as excel files, text files, images, audio, video).

Existing tools often focus narrowly on specific modes with lack of integration, leading to suboptimal decision-making processes and they do not fulfil with FAIR principles on their data management processes.

2. SIMULATION USING THE BFT

Secondly, depending on the final goal established by the user (as those set in MITHOS pilots), the platform will connect to different modelling / simulation tool that will feed the platform with high-added value data for the decision-making process. In particular, some of the simulation capacities that will be added during MITHOS project are the following: (i) evaluation of the attributes of road and cycling infrastructure with Lane Patrol; (ii) simulation of human behaviour / traffic models for bicycles and e-scooters with SUMO; (iii) real-time traffic situation assessment (includes traffic efficiency and safety) with ADASH, (iv) simulation of conflicts between modes (e.g. real time road-cycle collision warning) with MOTSS-BOB, OSTAR; (v) road Conditions Quick Assessment with RUCS; (vi) simulation of queuing times at charging stations with Eco-charging; (vii) emissions assessment with COPERT.

3. DATA PROCESSING IN THE FDM AND DATA SHARING STRATEGIES

Thirdly, the FSD module will guarantee the reliability of the information provided by the platform in terms of data interoperability, adaptability of the architecture, sustainability, security, accessibility, understandability, credibility and robustness, among others.

MITHOS will include a federated multi-layered dynamic database, which collects and combines different data from different tools/sources allowing real-time data availability and automated correlations and obtaining human- and/or machine-readable outputs. The data flow and tooling service pipelines will mainly focus on the connection of edge devices to feed the MITHOS main tooling system, aggregate and enrich them, and then continue flowing both – smart data and service operations – to the users’ desk for decision support. Also, connecting to the EMDS will enable MITHOS to add more datasets to its processing chain, improving the quality of the smart data.

On the one hand, all the heterogeneous data gathered are fed into the FSD module through logical (non-physical) standardized connectors based on IDSA approaches. These connectors will not only connect to data but also, they will automatically/dynamically provide/integrate a description and classification for each of the data, as well as managing and organizing their storage according to the type of data the connectors define (i.e. raw or smart data). This data management will be done by an ontology and metadata catalogue that will allow all the data to (i) be readable and understandable by both human and machine; (ii) ensure semantic consistency and interoperability of the data, and (iii) limit the probability of obtaining random correlations or false positives, increasing significatively the robustness and reliability of MITHOS platform. This metadata information will be relevant for data retrieval, access, quality control and improvement and will also solve the nature of the high heterogeneity of data structure. It will also facilitate the subsequent processing of the data (by the DSS module). All the information contained in Federated Data infrastructure will be naturally linked to the data catalogue. This FDM also deals with the phases of 8 | 45 MITHOS privacy and data protection inherently by design of connections between data sources and data users, as the data management platform does not necessarily contain data, but only links to location of specific data needed for a specific use case. Therefore, if there is no need for private data in the use case, no such data will be sourced or transferred through the data management platform.

On the other hand, the FSD module will feed the selected Fundamental Tools and will receive data back from the performed simulations and calculations for the measurement of key mobility or infrastructure service indicators (i.e., cost of infrastructure changes or maintenance actions, emissions from multimodal traffic, energy consumption in public transport management, safety of transport infrastructure users and workers, user accessibility to transport modes and need for EV charging points). As the Fundamental Tools are also connected to MITHOS platform by standardised IDSA connectors, the outputs of the simulations and calculations made are then managed and stored in the same way as data gathered from databases and sensors, that is, in this case, as smart data.

MITHOS will be also connected with the EMDS, currently under development and expected to be operational during the project time of MITHOS. EMDS might be built upon existing technologies and software stacks, e.g. EONA-X, Catena-X, so all of them are basically suitable to be connected to MITHOS from both – policy and technological state – thanks to the use of most commonly data space connector that will come from the open-source software stack of the Eclipse Data Space Components (EDC).

Finally, this module has an automated AE that will continuously monitor all new connections to MITHOS platform. When new data is connected, correlations between the data which is already available in the multi-layered database and the new data are analysed by the AE which finds new interconnections and put them into another layer of the database.

4. AI-BASED DECISION-MAKING
ANALYTICS

Finally, once all needed Smart Data has been processed and has become readable both by human and machines, this information is transferred to the DSS module through a universal and IDSA data (space) connector providing all aspects related to data sharing as specified in IDS-RAM18 and relevant standards for Data Spaces. The DSS module is composed of (i) a set of smart solvers, each addressing one specific objective of the infrastructure management considering a “multicriteria” analysis (e.g., safety and efficiency); and (ii) a high-level decision-making support tool, that will address complex infrastructure management problems combining solver’s tools. Interoperability and combination of the smart tools will be ensured by AI and/or imitation learning techniques to find a black-box representation of the relationships between inputs and outputs of the different smart tools.

On the one hand, within MITHOS, the DSS will include five solvers (that would be extended afterwards): (i) Design of public transport network; (ii) Design of last-mile IWW logistics; (iii) Design cycle lanes; (iv) Optimise location and sizing of charging infrastructure for EVs; (v) Predictive maintenance of road/cycling infrastructures. These solvers will have two fundamental characteristics: (i) be generic enough to be implemented and applied in any territory; and (ii) be interoperable, so they can be combined to face optimisation in more complex multimodal environments.

On the one hand, the high-level decision-making support tool will combine two or more smart solvers to find more meaningful trade-offs for the decision-makers, providing a comprehensive and easy-to-use DSS that will help them to make the "right" choice. In other words, it will be able to generate much more optimised and ambitious solutions for the redesign of infrastructure, based on multiple criteria and considering a multimodal network. Thanks to this innovative framework, transport authorities would address complex infrastructure management problems, such as: (i) Optimal investment plans for multimodal infrastructure topology modification re-design; (ii) Optimized infrastructure designs for enhanced multimodal transport offer; or (iii) Smart maintenance planning to reduce traffic disruptions and improve multimodal infrastructure resilience.

The optimal solutions provided by the high-level decision support tool will be shown through an open-source dashboarding platform that provides all necessary interaction and visualization (HMI) primitives, enabling both smart and high-level decision-making support tools and algorithms to be applied across various use-cases and contexts/environments, considering local constraints and decision variables. It will feature geographical representation for effectively representing inputs and appropriate visualization techniques for outputs. In this way, MITHOS platform grant users´ autonomy and allow dynamic interaction both with the DSS module and/or the Federated Data module. By incorporating the appropriate level of abstraction and design support features, such as scenario impact simulations and territorial visualization, the MITHOS platform will enhance interactive decision making among stakeholders and end-users.

5. ASSEESSMENT OF IMPACTS

Finally, the impact assessment module will include not only an evaluation at local scale but also will be expanded to global level by performing a Cost-Benefit Analysis (CBA) to assess the overall environmental and safety impacts. To do so, on the one hand, environmental benefits, compared to the baseline, will be determined using R-TAMS (IPFEN) at local level, and SIBYL and COPERT, at the global level. The results will include projections of air pollution, GHG emissions, and energy consumption up to 2050.

The analysis will also incorporate an infrastructure lifecycle perspective, utilizing an extension of SIBYL developed during MITHOS. On the other hand, safety assessment will be done with the Mobility Observation Box (MOB), an image-based sensor for measuring traffic conditions and traffic conflicts that enables to measure the safety of transport infrastructures according to objective criteria and thus make them comparable using ML algorithms and AI. This MITHOS module will calculate safety predictions and risk estimation based on changes of traffic volumes and changes in mode mixtures.