Summary
Presenting the different components of digital ecosystems with examples from emerging digital ecosystems research.
Ka mahia tenei rangahau ki nga whenua me nga wai o Te Kawerau a Maki me Ngāti Whātua Ōrākei me Ngāti Porou. Tēnei ka mihi! This research takes place on the lands and near the waters of Te Kawerau a Maki, Ngāti Whātua Ōrākei and Ngāti Porou. We acknowledge them!
We acknowledge that the concept of connections within place are a taonga from mātauranga Māori and are grateful to Dion Pou for sharing this knowledge from Māori.
The concept of digital ecosystems behind The Digital Ecosystems Project gave rise to further exploration of digital ecosystems in the Global Partnership of AI (GPAI)‘s Digital Ecosystems project. This research was published as a report for the GPAI Summit 2024. One of the key contributions from that report was a diagram of how the different components of a digital ecosystem fit together. This diagram is shown next and the rest of this blog will walk through the different components of the diagram with examples from emerging digital ecosystems research.
The digital ecosystems components are:
- Community
- Data gathering/generation
- Data structuring and storage
- Building understanding
- Exploring decisions
- Human machine interface/Communication space
I’ll summarise each of these components next – including examples – and finish with an overall summary of digital ecosystems in practice.
Community
The key motivation for digital ecosystems was to democratise digital technology so that a group of people with a shared interest, broadly referred to as a community, could gain a shared understanding of an issue/system and problem solve collectively. This community can consist of several different sub-groups depending on the system being considered. One example is a digital ecosystem for providing better healthcare in a hospital environment. In that case the community would consist of patients (that need care) and clinicians (that provide care), but also administrators that organise the people & resources required for care, perhaps suppliers that provide the equipment & supplies required for care, and the whānau or (extended) family of patients that need to both understand the care pathway their loved one is undertaking and be available to support them during that pathway, e.g., with visits and/or to bring them to and from the hospital. Another example is a digital ecosystem for supporting katiakitanga | conservation/guardianship of the taiao | environment. In that case the community would consist of people who live on or near the taiao under consideration, people with a deep connection to the taiao, e.g., indigenous people from that taiao, local/regional/national/global governments that contribute to the management of the taiao, scientists and citizen scientists that work (either remotely on in person) on the taiao, and possibly future generations of people that will be affected by the ongoing condition of the taiao.
Data gathering/generation
When using digital technologies to better understand and/or problem solve about a system, gaining knowledge through data is critical. In some cases, data will already be available, but may be distributed or fractured and require gathering together. In other cases, data may not be available and digital technology, such as Internet-of-Things (IoT) devices may be required to generate (i.e., capture and transmit) data. In the hospital example, data will be available via patient health records, staff rosters, hospital layout, etc but useful other data can be generated, e.g., via automatic timestamping technology that captures a patient journey as they move through a care pathway. In the example of kaitiakitanga taiao | environmental conversation, existing data on the taiao | environment may have already exist through a variety of environmental monitoring initiatives, but can be complemented by IoT sensors for predator control and water quality, video capture (e.g., from 360 degree cameras) of the taiao, etc.
Data structuring and storage
Streams or channels of data are important, but without context the data can be opaque. For example, (-36.853909, 174.645269, 17) could be the depth in mm of a stream or the temperature in Celsius at a given latitude, longitude coordinate. By adding data ontologies, i.e., structured representations of data concepts and relationships to enable interoperability and integration [Data Ontology], multiple streams of data can be coalesced into one or more data stores, i.e., platforms for storing structured, for use within digital models (see Building understanding). In the hospital example, data can be timestamped data about a patient journey, e.g., time in/out of surgery, or activities of staff, e.g., pick ups and drop offs by transit nurses, or might be data for the patient’s health record, e.g., blood pressure, temperature during obs. The example that started this section, i.e., (-36.853909, 174.645269, 17), demonstrates the kind of data from the example of kaitiakitanga taiao | environmental conversation. In that example data will often be geocoded, i.e., with a latitude, longitude, sometimes an elevation, but will refer to a specific measurement, e.g., water depth, temperature. In some cases, the data will belong to a specific “point of interest” or device, e.g., pest traps will often have time series of number of pests caught associated with them. In all these examples, the data ontology is key for understanding what the data means and enabling unification of data sources.
Building understanding
At the core of a digital ecosystem is a digital model, e.g., a digital twin, that takes as input the unified data from the system and uses this input within an in silico environment (i.e., digital, on a computer) with the model producing output of various system metrics and/or visualisations. The model represents the system of interest to the partner community and its output help that community to build an understanding of that system across various facets and over time (i.e., longitudinally). For the hospital example the model may be a realistic 3D simulation of the hospital or may be a logical simulation of patient/clinician flows in the hospital. The metrics could be the occupancy/busyness of various departments as well as the workload of the clinicians. Getting a better understanding of how these change will help decision making in terms of hospital management. The model for the example of kaitiakitanga taiao | environmental likely be a geospatial model with data annotated to key locations and measuring things like water quality, wildlife abundance, pests eradicated, etc. Getting an understanding of how these change over time and any influence one may have on the other, e.g., how eradicating pests helps improve wildlife abundance, is key knowledge for effective kaitiakitanga.
Exploring decisions
Building a community’s understanding of the complex system they share an interest – with data they may have contributed to – may be sufficient in some cases. In many cases, however, the community will want to explore decisions to manage and/or improve various metrics of the system of interest. For the katiakitanga taiao example, a community might want to determine how to distribute pest control technology to reduce pest abundance and revitalise native flora and fauna. In order to explore decisions, scenarios and policies, the digital model at the heart of the digital ecosystems much be augmented by one or more AI models. These models predict how the system changes over time when a given decision, scenario, policy is put in place. Note that many of the same interface/communication (see Human machine interface/Communication space) approaches use with the Building understanding component will also be useful for Exploring decisions, but extra knowledge, methods, technology, etc is applicable too. For example, suggestion engines and/or serious games can empower a community to “play” with decisions and problem solve in an innovative, inclusive, creative fashion. For the hospital example, managers could experiment with different staffing level and patient priorities to make sure their hospital is working as smoothly as possible given desired/required patient outcomes and staff workloads.
Human machine interface/Communication space
The digital technology embedded within a digital ecosystem – from the IoT sensors through the digital model of the system of interest and including the AI approaches for predicting how that system behave – are all working to help a community problem solve/decision make for their complex system. The interface between that community and the rest of their digital ecosystem happens via a Human machine interface which creates a Communication space that people in the community can use to communicate with the model(s) of their system and with each other about that system. The Communication space represents the safe exploration space at the junction of the community and the Human machine interface(s) where community members can debate while they explore data-informed decisions until they reach agreement(s). A variety of technology can be used to create the human machine interface(s) and, hence, the Communication space such as electronic dashboards, virtual and/or augmented reality, and serous games. For the hospital example, dashboards are already common in hospitals so presenting the outcomes of the digital model as dashboards within a serious game would enable hospital management to creatively problem solve issues such as emergency department (ED) surge. Virtual reality walkthroughs of the taiao in which the condition of the taiao is shown in a time-lapse manner is an engaging and informative interface for exploring decisions for kaitiakitanga.
Summary
Digital ecosystems aim to democratise digital technology such as IoT Sensors, data stores, digital models, e.g., digital twins, and AI so that groups of people (communities) with a shared interest can creatively problem solve and collaboratively make decisions. In this post we explored a little the components of a digital ecosystem and what these components might look like in a hospital setting and when helping a community with katiakitanga taiao | environmental guardianship. By empowering communities with digital technology, the benefits of creativity, inclusivity and diversity can help those communities together to explore the everyday decisions they face about their complex systems.