Summary on integration of solutions
Designing the best digital solution might require the implementation of different tools as a set of measures. Therefore, the subsequent table summarizes the potential integration and gives examples.
Artificial Intelligence (AI) | Data Sources | Digital Twins | Communication and collaboration tools | Earth Observation/ Geospatial Tools | Mobile tools | (Remote) Management Information Systems (R/MIS) | Internet of Things (IoT) | |
---|---|---|---|---|---|---|---|---|
Artificial Intelligence (AI) | ■ | ■ | ■ | ■ | ■ | ■ | ■ | ■ |
Data Sources | Both Big Data and Open Data function as inputs for ML and DL models. Big data might include text data from social media, where large language models (LLM) or natural language processing models (NLP) models might be useful. Open Data could also include datasets like OpenStreetMap, which can be used and analysed by both types of models. An example is the use of Big Data collected from multiple river sensors to train a neural network for flood prediction (Sit and Demir, 2019) | ■ | ■ | ■ | ■ | ■ | ■ | ■ |
Digital Twins | AI enables to perform simulations within digital twins, allowing for predictions of future states and process optimisations that can be applied in corresponding physical systems. The decision making-capability of AI can be leveraged to automate urban systems. In a digital twin for Copenhagen, machine learning is being used to support sea-level rise scenario simulations, for example (Henriksen et al. 2023). | Big data technology harmonizes multiple data sources, handling great volumes of diverse formats. Digital twins can use Big Data as input for more comprehensive and accurate representations of physical systems. Given the high velocity of Big Data updates, they enable actualized simulations and monitoring within the digital twin. This is the case with Pipedream, a digital twin for natural and urban drainage systems (Bartos and Kerkez, 2021). It integrates big volumes of data collected from sensors in a watershed to manage the system in real-time. | ■ | ■ | ■ | ■ | ■ | ■ |
Communication and collaboration tools | AI outputs can be integrated with the communication channels of a city administration and even automate them. Similarly, e-learning tools can benefit from AI through customization of learning, production and selection of new learning materials. A practical example of combining AI with communication channels is the use of ML to process social media data for identifying emergencies (Imran et al., 2015). | Data stemming from communication channels can function as input for Big Data systems. As an example, data generated on usage patterns in e-learning platforms could be integrated with other datasets to personalize and enhance learning. The integration of Big Data with communication tools is most evident when social media data is used for further analysis (Imran et al., 2015). | Digital twins and communication tools enable collaborative decision-making. Scenarios simulated can be shared easily and be discussed to inform decision-making. In the context of e-learning, digital twins can support scenario-based learning. Stakeholders can take advantage of these simulated scenarios and train for specific events (Therias and rafiee, 2023) | ■ | ■ | ■ | ■ | ■ |
Earth Observation/ Geospatial Tools | Earth observation tools make intensive use of AI models and methods. For instance, convolutional neural networks (CNN), a DL model, is extensively used when analysing satellite imagery. Furthermore, earth observation tools can function as data inputs for AI systems, supporting urban analytics processes and data-driven decision making. Although not in an urban context, the integration between remote sensing data, surveys and ML is used to predict floods in Bangladesh (Patel ,2024). | When only being used for data production, earth observation tools can be integrated into Big Data systems through harmonization with pre-existing information. , Due to the nature of earth observation data being space and time-specific, the frequency and coverage of data is key to obtaining more representative and relevant datasets. Moreover, the data can be shared as Open Data, increasing its utility and accessibility. The integration of Open Data with earth observation tools is exemplified by the monitoring system of severe droughts in Cape Town, South Africa (Van Belle & HIabano, 2019). Among other strategies, GIS tools were used to map areas with shortages and those with current water access. These maps were made public, supporting better resource allocation and community awareness. | BIM or Digital Twins, and GIS or Remote Sensing can be merged to visually map the physical infrastructure and the integration of geospatial data, such as land use and topography. An example for integrating MIS and Remote Sensing can be found in drought monitoring and prediction systems. Such systems make use of updated remote sensing data to track droughts and train models for prediction (Hao et al., 2014) | As with other technologies, Earth observation tools are supported by Communication channels and tools for dissemination, joint work and decision-making. For example, remote sensing data on phenomena such as heating islands or carbon emissions can be communicated effectively to citizens and stakeholders in real-time. Knowledge on the use of GIS and RS can be obtained through e-learning. | ■ | ■ | ■ | ■ |
Mobile tools | Mobile tools can be integrated with AI to automatize operations and customize the user experience. Algorithms can analyse data collected via mobile tools for predictive modelling, anomaly detection, and pattern recognition. Additionally, crowdsourced data can be analysed in real-time to provide immediate insights. When used as data sources, mobile tools can function as input for AI models. Such integration can be seen when processing mass social media data from mobile tools to identify emergencies in real time (Imran et al., 2015) | Due to mobile tools’ data collection nature, big data technology is frequently used in storing and harmonizing the collected data. Generated data can also be made available as open data. An example for the integration of Mobile tools and Open Data is the online application Gieß den Kiez of CityLab Berlin (n. d.). The platform combines crowdsourced data on tree watering in Berlin with an open data portal. | Mobile devices can interact with digital twins to provide data or to generate visualizations of specific data coverage patterns | Mobile tools can facilitate communication and collaboration, enabling distant and emergency communication, as well as e-learning. Mobile tools not only enable e-learning but also offer the opportunity for more personalized formats and can be the subject of e-learning. | Mobile devices can collect geospatial data and inform geospatial models. The portability of mobile tools enables users to cover remote areas. The integration of data from earth observation tools with that collected from mobile devices allows to formulate maps and understand the spatial distribution of data, especially in relation to the built environment. In the context of floods, the combination of both technologies can be of great support when working with a centralized system. Mobile tools can supply real-time data about the geographic dispersion of people during such events, while earth observation tools can provide broader context (Yuan et al., 2021) | ■ | ■ | ■ |
(Remote) Management Information Systems (R/MIS) | AI can support systems to identify patterns and make forecasts, enabling to learn from historical data and generate more accurate analytics. Advanced MIS can incorporate artificial intelligence to process data and provide insights or automate processes of decision-making. An example for this integration is the Global integrated drought monitoring and prediction system (GIDMaPS) (Hao et al. 2014). It makes use of a neural network in conjunction with a Bayesian algorithm for predictions. | The integration of Big Data and MIS improves the capacity to provide accurate insights, since the data is more varied and covers diverse areas and formats. Additionally, open datasets are a relevant alternative for acquiring and feeding data into the system, while insights or results from MIS can be shared publicly as Open Data. Rio de Janeiro's Centre of Operations (COR) provides an example in the integration of MIS technology and Open Data. This operation center makes public its everyday management data, including weather and crime for instance (Cosgrave et al., 2013) | MIS can help in designing and managing infrastructure or urban layouts that are more resilient to climate and disaster risk impacts. A concrete integration of both technologies are digital twins specifically designed to support situational awareness (Fan et al. 2019). | MIS can serve as a way to share information and coordinate efforts among different stakeholders. E-learning on MIS can be used to build needed capacities. | Spatial data is key for the use of MIS in disaster risk management. It allows for geo-localized visualizations of data, which helps in operation during crisis or mapping vulnerabilities. Remote sensing can also be relevant to provide relevant spatial data to the system. A good example for this is wildfire prevention and management. Remote sensing data is continuously monitored and integrated into MIS, supporting early detection (Kalabokidis et al., 2016). | Mobile tools can provide real-time data for MIS, thereby enhancing decision-making processes. Given the flexibility of mobile devices, such as smartphones, their data can more accurately reflect dynamic scenarios. Furthermore, MIS can be displayed in mobile devices, providing access to information to a wider audience. Integration between the two technologies is conceivable in efforts to monitor social media during disasters, such as in cases of floods (Arthur et al., 2018). | ■ | ■ |
Internet of Things (IoT) | The Internet of Things (IoT) functions as data sources for AI algorithms. With wide coverage, IoT enables comprehensive urban planning powered by AI. Additionally, AI also shapes the IoT landscape, especially through the Artificial Intelligence of Things (AIoT), a more recent class of interconnected devices. In the AIoT, an AI algorithm makes decisions on multiple devices based on the information gathered. The Urban Systems Lab (n. d.) provides an example of how to integrate remote sensing data with IoT to produce better data for AI training towards the use of Urban NbS. | Given IoT’s data collection nature, it can consistently cover specific areas and domains in a controlled fashion. When integrated with Big Data for storage, processing and accessibility, it can be made available as input for other technologies. Additionally, like other data collection tools, the data generated by IoT can be made public as Open Data. An example for this integration is Colombia’s National Unit for Disaster Risk Management early warning system. This system describes the process of transforming raw data to usable one, is the Early Warning (Libelium, 2021). It collects data from IoT sensors and processes and aggregates it for corresponding use. | IoT sensors, such as. RFID and laser scanners, collect real-time data of the physical environment. They can provide a constant flow of information to digital twins, enabling an accurate representation of the physical urban system. Examples for this interconnection are varied, as IoT sensors compose one of the backbones of digital twins. Multiple cases demonstrate the integration for early warning systems and general management (Riaz et al. 2023). | IoT can serve as a device for registering information, generating alerts and notifications to be communicated to citizens. Communication can also be further enhanced through IoT data made public. | Earth observation tools and the IoT can work as complementary data collection systems. IoT collects granular real-time events, while earth observation tools capture broader-scale data, although less frequently. Furthermore, IoT might be able to support model calibration whenever geospatial models are being trained. A good example for this integration can be found in real-time GIS dashboards. These dashboards display geo-localized developments as those registered by IoT sensors in real-time (Lewis, n.d.) | Mobile tools can complement sensors by collecting real-time data. Together, these technologies enable the collection and integration of data into urban analytical models for deeper analysis. A concrete example of how both technologies can work together is in flood risk monitoring and situational awareness. IoT sensors and mobile tools can simultaneously provide data to a centralized system (Yuan et al., 2021). | The integration of IoT wireless sensors, including cameras and drones, can provide near-real-time data to MIS, which can be used for not only planning but also monitoring interventions. A good example on the integration of IoT and MIS is in landslide-specific early warning systems (Kuradusenge et al., 2022). Here, soil moisture sensors are integrated with a MIS to generate timely warnings. | ■ |
Licence
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0).