Home Articles Standards: Geodata Fusion Study Shows Value of Open Standards

Standards: Geodata Fusion Study Shows Value of Open Standards

<< In an age of information overflow, standards are essential to ensure communication between various applications and sharing of data from multiple sources. Open Geospatial Consortium (OGC) has been actively working with industry and security agencies to develop open and international geospatial standards >>

In the 2010 Open Geospatial Consortium (OGC) Fusion Standards study, Data Fusion was defined as: “the act or process of combining or associating data or information regarding one or more entities considered in an explicit or implicit knowledge framework to improve one’s capability (or provide a new capability) for detection, identification or characterization of that entity.” Though the focus of the study was on military intelligence (INT), decision fusion is relevant to business intelligence, urban planning and many other domains.

Interoperability based on open standards is radically changing the classical domains of data fusion while inventing entirely new ways to discern relationships in data with little structure. Associations based on locations and times are of the most primary type. For OGC, ‘Open Standard’ means that the standards document is freely and publicly available in a non-discriminatory fashion with no license fee, vendor and data neutral, and agreed to by a formal consensus process.

Figure 1:Two examples of Fusion based on OGC WMS standard. Each composite web map view shows several distinct layers of geospatial information fused into a single map visualisation

Recommendations from the study were subsequently addressed in the OGC series of OGC Web Services (OWS) testbeds including OWS-8 in 2011 and OWS-9 in 2012. Results of the testbeds were recorded in OGC Engineering Reports[1] and videos[2], demonstrated to the sponsors and additional communities, and have affected further development of OGC standards.

Geodata Fusion and Open Standards
Many of the fusion processes described here can be achieved in closed architectures with existing, single-provider software and hardware solutions. However, without the use of open standards multiple islands of data and services emerge that are difficult to automate and scale. Standardsbased data, applications and services enable an automated and interoperable fusion environment supporting secure sharing of data and transparent reuse of ‘pluggable’ services for handling large data volumes and unanticipated analytical challenges.

Some elements of the desired open standards-based fusion framework are pervasive now. For example, the OGC Web Map Service (WMS) enables fusion of maps. WMS allows for maps as pictorial layers from different sources to be geographically overlaid to create a composite map suitable to the user’s need (Figure 1).

Three categories of data fusion (Figure 2) were used to organise the study: Observation Fusion; Feature Fusion; and Decision Fusion. These categories are described below.

Figure 2: The three data fusion categories used in this study: with sources of geodata and an increasing semantic context

Observation Fusion
It involves merging multiple sensor measurements of the same phenomena into a combined observation. Fusion processes include combination of various sensor measurements into well characterised observations including uncertainties, for example, to support signature analysis.

The basic requirements for sensor fusion include:

  • Discovery of sensor systems, observations and observation processes that meet a user’s immediate needs
  • Determination of a sensor’s capabilities and quality of measurements
  • Access to sensor parameters that automatically allow software to process and geo-locate observations
  • Retrieval of real-time or time-series observations in standard encodings including encoding the uncertainty of the measurement, and parameters need to process the measurements
  • Tasking of sensors to acquire observations of interest
  • Subscription to and publishing of alerts to be issued by sensors or sensor services based upon certain criteria
  • Entity identification, classification and association
  • Enablement of fusion processing by providing access to processing engines and needed reference information (for example, signatures and training data)

Much information suitable for fusion begins with or is derived from observations by sensors or humans. This is particularly true for information that is highly dynamic in nature and of a timely nature. These observations, either raw or processed, can serve as input into fusion processes or they may be used to identify recognisable objects that are then treated as input into a fusion process.

Standards for Observation Fusion are relatively mature; in particular the OGC Sensor Web Enablement (SWE) standards have been adopted as consensus standards with implementations for several years. The SWE architecture document provides an overview[3].

A study on SWE Implementation Maturity is currently underway. The study is considering implementation of SWE by US DOD, NASA, NOAA, EEA, and many other programmes.

The following areas for future work were developed in the Fusions Standards study for Observation Fusion

  • Coverage fusion based on Web Coverage Service and Web Coverage Processing Language.
  • Further develop Event handling in the OWS Architecture
  • Use of open standards for Motion Imagery and location – coordinated with MISB
  • Apply SWE to Mobile Internet and opportunistic sensing
  • Further develop Secure Sensor Web by applying security services to SWE
  • Registries for Sensor/ Observation Fusion, for example, for signatures
  • Online community sanctioned definitions for sensor terms
  • Harmonisation of the process of precise geolocation
  • Characterising and propagating uncertainty of measurements
  • Increasing use of geometric and electromagnetic signatures

Figure 3: Feature Fusion workflow example from the OGC Web Services, Phase 5 (OWS-5) testbed. The workflow is composed of a conflation service and a topology quality assessment service operating on features.

Feature Fusion
Feature fusion includes processing of observations into higher order semantic features and feature processing. It improves understanding of the operational situation and assessment of potential threats and impacts to identify, classify, associate and aggregate entities of interest. Feature fusion processes include generalisation and conflation of features. Conflation technology offers useful options to deal with imperfect, heterogeneous, conflicting and duplicated data.

A service-oriented architecture is well suited to support distributed conflation rules services. Feature fusion yields information resources that are more powerful, flexible and accurate than any of the original sources. A workflow of Feature Fusion is presented in Figure 3 including conflation rules and processing services followed by a Topology Quality Assessment (TQAS) service.

Recommendations from the Fusion Study Phase 1 were implemented in the OGC Web Services, Phase 7 (OWS-7) testbed. OWS-7 built on the OWS-6 Geoprocessing Workflow and Decision Support Services work. It employed fusion for Feature and Statistical Analysis (FSA). These recommendations relevant to Fusion Standards Study were noted:

  • WPS profiles are needed in order to achieve semantic interoperability of geoprocessing
  • Designing WPS Profiles is a challenge not only regarding choosing the appropriate input and output type definitions, but also regarding choosing appropriate classifications
  • Metadata profiles for registering WPS in OGC Catalogues are missing and hinder the use of the publish-find-bind pattern

Decision Fusion
Decision fusion involves processes supporting a human’s ability to make a decision by providing an environment of interoperable network services for situation assessment, impact assessment and decision support, using information from multiple sensors and processed information.

Decision Fusion provides analysts an environment where they can – using a single client interface – access interoperable tools to review, process and exploit multiple types of data or products from multiple sensors and databases. Decision Fusion includes the use of information from multiple communities, for example, multi- INT, in order to assess a situation, and to collaborate with a common operational picture. This study also considered more recent advances such as social networking to support decision fusion.

Figure 4 shows an example of decision fusion that goes beyond current web mapping tools to associate trends and causes from multiple sources. In an Afghanistan election attack scenario (Figure 4), a likely target for an attack against the Pashtun during the 2009 election was considered to be in Jalalabad, a largely Pashtun area with strong ties to Karzai, and a target of IED. The largest polling station in the area was at Compano Mosq, with an estimated 13,700 voters. An attack here on an election day was likely to impact the outcome.

The ‘Decision Fusion Node’ defined in this study is a scalable concept ranging from a person with a mobile computer to a Fusion Center such as the Information Sharing Environment (ISE) Fusion Centers as operated by the US Department of Homeland Security[4]. Implicit in the concept of the Decision Fusion Node is the collaboration with other nodes, for example, distributed decision fusion.

Information available to an operations node is from multiple intelligence collection types (multi- INT). Intelligence sources are people, documents, equipment or technical sensors, and can be grouped according to intelligence disciplines:

  • Human intelligence (HUMINT)
  • Geospatial intelligence (GEOINT), including Imagery Intelligence
  • Signals intelligence (SIGINT)
  • Measurement and signature intelligence (MASINT)
  • Open-source intelligence (OSINT) (OSINT involves finding, selecting and acquiring information from publicly available sources, as opposed to covert or classified sources; it is not related to open-source software or public intelligence.)
  • Technical intelligence (TECHINT)

Examples of multi-INT for an urban situation are shown in Figure 5[5]. Information elements are placed in the table according to generating source (header row) and classification between hard and soft information (below and above the diagonal, respectively).

A conclusion of the Decision Fusion Workshop is that additional development remains to be done on standards for structured information that supports Decision Fusion both internal to OGC and with other standards bodies. For example, further work is needed on methods for schema mapping, like identification of rules for mappings, and an increase in the focus on handling of associations as the identification of an association between entities is at the heart of fusion. Developing a ‘Decision’ as first class object in information modelling will assist in quick response in template form. The most effective environment for accomplishing the various types of fusion is expected to be a networkcentric architecture with distributed databases and services based on a common core of standardsbased data formats, algorithms, services, and applications. Such an environment allows the various forms of information to be collected, stored, managed, fused and disseminated vertically (from international to individual level) and horizontally (peer to peer).

Fig 4: Decision Fusion Ex (Source: FortiusOne)

The Fusion Study identified a framework for advancements in observation fusion and feature fusion. Decision fusion recommendations have been addressed but much more can be done in this area. Open standards about decision fusion are still needed, for example, to support the pattern of ‘if this event occurs, then consider these actions.’ For further information, see the OGC Fusion Standards Study, Phase 2 Engineering Report[6] and the OWS- 8 Demonstrations webpage[2].

Further work on fusion should consider the great advances in mobile development. Mobile devices now serve as decision fusion platforms and they are increasingly loaded with sensors that can be used for fusion local to the device or part of the larger sensor web. The Internet of Things approach opens a broad set of fusion topics that can be addressed for multiple domains.

Figure 5 – Multi-INT examples for an urban situation


  • OGC Engineering Reports https://www.opengeospatial.org/ standards/per
  • OGC Web Services Testbed, Phase 8 (OWS-8) Demonstrations, https:// www.opengeospatial.org/pub/ www/ows8/index.html
  • Botts, M. et. al., “OGC Sensor Web Enablement (SWE): Overview And High Level Architecture, ” OGC White Paper, OGC Document OGC 07-165r1, (2 April 2013) . Fig 4: Decision Fusion Ex (Source: FortiusOne) https://portal.opengeospatial.org/ Figure 5 – Multi-INT examples for an urban situation files/?artifact_id=48492
  • Global Justice Information Sharing Initiative (Global). “Baseline Capabilities for State and Major Urban Area Fusion Centers: A Supplement to the Fusion Center Guidelines” US Government, Department of Justice (2008).
  • Pravia, M., “Generation of a Fundamental Data Set for Hard/ Soft Information Fusion”. Fusion 2008: The 11th International Conference on Information Fusion. Cologne: International Society of Information Fusion (2008).
  • Percivall, G., ed., “OGC Fusion Standards Study, Phase 2 Engineering Report,” OGC Document 10-184, (13 December 2010). . org/files/?artifact_id=41573