Below are some of the projects I have worked on. More information about each of the projects can be found on the School's project pages.
The Virtual Vehicles project is based on the work of the Italian neuroscientist and cyberneticist, Valentino Braitenberg. Through a series of elegant thought experiments, published in his 1984 book, Vehicles: Experiments in Synthetic Psychology, Braitenberg suggested that surprisingly complex behaviour could emerge from the interaction between forces and factors that were distributed across brain, body and world. In particular, once we factor in the contributions of the physical environment and bodily structure, the kind of behavioural control mechanisms required to yield intelligent behaviour may, Braitenberg suggested, be quite simple. The Virtual Vehicles project is an attempt to extend Braitenberg’s original work using state-of-the-art approaches to the creation of virtual robotic platforms and interactive 3D environments.
Virtual Vehicles uses the Unity game engine to construct a series of simple four-wheeled vehicles that are embedded in a virtual 3D environment. The behaviour of each vehicle is controlled by a library of custom Unity components that act as the vehicle’s nervous system. Vehicles can perceive their environment using visual, olfactory and tactile information. Visual information is provided by camera components that function as the vehicle’s eyes. They render the current scene (from the vehicle’s perspective) to a 2D texture asset that is then subjected to image processing. Tactile information is provided by sensors attached to hinge joints or C# routines that rely on RayCasting techniques. Motor output takes the form of physical forces that are applied to Wheel Collider components. These change the angular orientation and rotation of the vehicle’s wheels and control the vehicle’s movement in the 3D landscape.
The ITA ACT-R CSSC project forms part of the International Technology Alliance (ITA) research programme. The ITA ACT-R CSSC project aims to develop a Cognitive Social Simulation Capability (CSSC) based on the ACT-R cognitive architecture. Cognitive social simulation is a computer simulation technique that attempts to understand the interaction between cognitive, social and technological factors in a variety of different task contexts. This makes the technique useful for understanding situations in which multiple individuals must work together to perform a cognitive task. Some examples include a team of medical practitioners working together to diagnose an illness, a team of intelligence analysts assessing intelligence information, and a team of engineers working to design a new product. All these tasks are ones in which aspects of human cognition play an important role. However, they are also ones in which issues of social engagement and interaction are of critical importance. The ITA ACT-R CSSC project aims to support computer simulations that emphasize the potential interactions between cognitive, social and technological factors in team-based tasks. The ACT-R CSSC simulation capability relies on the use of the ACT-R cognitive architecture to model the capabilities and constraints of the human cognitive system in a realistic manner. This enables the simulations to accommodate mechanisms for learning, memory, attention and reasoning. In addition, the ACT-R CSSC provides support for the modelling of social processes. These allow individuals to communicate and share information, thereby influencing the activity of other team members.
The IEXTREME project was a trans-Atlantic collaborative project, funded by the U.S. Office of Naval Research. The project was a collaborative venture between the University of Southampton, Applied Research Associates and Rababy & Associates LLC, with Applied Research Associates acting as the prime contractor. The main goal of the IEXTREME project was to develop a better understanding of the ideological enablers associated with the behaviour of terrorist and insurgent groups. The National Military Strategic Plan for the War on Terrorism identifies extremist ideology as the enemy’s strategic center of gravity, and the Department of Defense (DoD) plays a significant role in establishing an environment unfavourable to extremist ideas, terrorist recruitment, and support. In spite of this, however, we have, as yet, little understanding of the specific ways in which extremist ideology contributes to various forms of terrorist action. IEXTREME aimed to address this shortcoming by combining state-of-the-art approaches to cultural modelling with a variety of advanced knowledge technologies. The project built on the scientific and technical outcomes of a number of previous projects, including SEMIOTIKS, MIMEX, ITA, ArtEquAKT and AKT.
The ITA CENES (Cognitive Extension in Network EnvironmentS) project formed part of the International Technology Alliance (ITA) research programme. ITA CENES aimed to explore a number of issues related to the potential impact of contemporary and near-future network environments on human cognitive processing.
One line of research in the ITA CENES project explored the effect that network-level variables (such as network topology) have on collective (group-level) problem-solving performances. A number of studies have shown that, at least in some information processing contexts, certain kinds of network (e.g., small-world networks) may be better suited to supporting collective problem-solving than others. Our aim in the CENES project was to examine these kinds of phenomena in order to better understand the socio-cognitive implications of specific network types.
Another strand of research in the ITA CENES project concerned the notion of what we broadly refer to as the 'network-extended mind'. In attempting to understand human cognition, cognitive science has tended to focus on the brain as the sole mechanistic substrate of mental phenomena. Cognitive processes, as well as the familiar elements of mentalistic discourse – the mental states that we use to explain and predict human behaviour – are typically seen as arising solely from the operation of the biological brain, and the physical machinery of the human mind is typically seen as something that is solely located inside the human head. In contrast to this rather neurocentric view, the extended cognition or extended mind perspective argues that the physical mechanisms of the human mind are, at least occasionally, distributed, or extended, in nature. According to this perspective, which has been championed by the philosopher Professor Andy Clark, the brain should not be seen as the sole point source of mental phenomena. Rather, the brain should be seen for what it really is: one element of a complex extended network whose representational and computational components include aspects of the extra-organismic social, cultural, linguistic and technological environment. It is from the dynamic interaction of these various resources (in conjunction with the biological brain) that many aspects of the human mind are, we suggest, materially constituted. This is clearly a bold claim, but is it a correct claim? Our aim in the ITA CENES project was to understand how networks (of many kinds) contribute to the modification, and indeed realization, of human cognitive states and processes. We were particularly interested in the way in which our increasing access to a range of evermore sophisticated network-enabled devices and network-accessible information resources acted to transform our traditional notions of ourselves as biologically-bounded cognitive agents.
A final area of research activity in the ITA CENES project was the notion of shared understanding. Conventional views of understanding typically make reference to a range of psychological states and processes that support a degree of predictive or explanatory competence in a particular domain of discourse (see Smart et al., 2009). Shared understanding on this view is typically regarded as the possession of equivalent mental models or domain-relevant knowledge. In contrast to this view (but not necessarily in opposition to it), we suggest that the mechanisms that realize shared understanding are not necessarily restricted to inner (in-the-head) processes. Instead, such mechanisms can extend beyond the boundaries of skin and skull to incorporate a variety of environmentally-extended processing routines. In essence, we propose that shared understanding is something that can be examined and explained by externalist accounts of cognition, and that such accounts complement (but do not necessarily compete with) more conventional approaches to shared understanding (e.g. those being explored by other projects in the ITA programme). Our approach to shared understanding is thus clearly one that embraces a distributed approach to human cognition, and, interestingly enough, this aligns itself with recent efforts to understand yet another human factors construct, namely the construct of situation awareness (see Salmon et al, 2009). Since we have suggested that situation awareness is a particular form of understanding, namely dynamic situational understanding (see Smart et al., 2009), a distributed approach to situation awareness is largely compatible with the approach we adopted in the ITA CENES project.
SEMIOTIKS was a 3-year collaborative project between the University of Southampton and QinetiQ that was undertaken as part of the Data and Information Fusion Defence Technology Centre initiative. SEMIOTIKS aimed to address some of the challenges faced by military and civilian agencies in leveraging the potential of large-scale information networks to support enhanced situation awareness and information superiority. The project combined state-of-the-art approaches to text analysis and resource classification with semantically-enriched approaches to information retrieval, knowledge extraction, knowledge discovery, text summarization and knowledge dissemination. A key aim of the project was to facilitate the identification, classification and processing of unstructured textual resources by capitalizing on the availability of natural language processing (NLP) and machine learning (ML) techniques. NLP technologies were used to extract relational information from information resources in a form that was suitable for knowledge processing, while ML techniques supported the classification of resources with respect to the elements of domain ontologies. For demonstration purposes, SEMIOTIKS focused on two problem domains: the domain of humanitarian demining and the domain of improvised explosive device disposal. Within these two domains, SEMIOTIKS technologies were used to support the classification of domain-relevant resources (e.g., incident reports and humanitarian demining survey reports), the extraction of key factual information from unstructured and semi-structured textual reports, and the publishing of factual information content in linked data formats.
In addition to the focus on semantically-mediated approaches to information extraction and resource classification, SEMIOTIKS also sought to build on the user interface design work that was initiated in the AKTiveSA project. As a result of this work, SEMIOTIKS delivered a variety of user interaction technologies that were integrated into a Web-based Technology Demonstrator System (TDS). Some screenshots of the SEMIOTIKS TDS can be viewed by clicking on the thumbnail images below.
The MIMEX project was a Data and Information Fusion Defence Technology Centre (DIF DTC) cluster project comprising two academic partners (Cardiff University and the University of Southampton) and one industrial partner (General Dynamics UK Ltd). MIMEX aimed to investigate some of the key challenges confronting military agencies in an era of effects-based operations and network-enabled capabilities. These challenges include the need to integrate information from physically disparate and semantically heterogeneous information repositories, the need to coordinate response outcomes with bodies of socio-cultural and psychological information, and the need to cope with hostile agencies that may deliberately attempt to subvert or disrupt coalition decision making. To address these challenges MIMEX focused on a solution strategy that embraced semantic integration techniques, human factors analysis, trust evaluation and cultural modelling. The operational focus area for MIMEX was Stability and Support Operations (SASO), specifically Information Operations (IO). IO forms part of a spectrum of military operations that entail the exploitation of Open Source Intelligence (OSINT), often focused on the civil, rather than the military, domain. Not only are these operations a central focus of effects-based operations, they also reflect much of the operational reality of current military engagements by British Armed Forces. While traditional warfighting operations typically target effects against an enemy's ability to fight, operations such as IO often attempt to achieve regional stability by targeting the 'hearts and minds' of relevant social groups (i.e., those groups with sufficient power and influence to instigate or control negative events). This focuses attention on the need for enhanced cultural awareness, which was a key aspect of the ontology engineering effort for MIMEX. The aim was to develop ontologies that supported the ability of military coalition forces to engage in culture-sensitive decision-making (i.e., decision-making that was sensitive to the cultural idiosyncrasies of a specific area of operations). This focus on cultural issues served as the basis for projects such as the IEXTREME project, which attempted to combine Web-based technologies with cultural modelling techniques in order to support our awareness and understanding of culturally-disparate groups.
AKTiveSA was a 3 year project that was undertaken as part of the UK's Data and Information Fusion Defence Technology Centre (DIF DTC) initiative. The principal aim of the AKTiveSA project was to explore how advanced knowledge technologies could be used to support enhanced situation awareness and inter-agency collaboration in a specific operational context, namely humanitarian assistance and disaster relief (HADR) operations. As highlighted by the MoD's Joint Warfare Publication on Humanitarian/Disaster Relief Operations (JWP 3-52), HADR operations often require military forces to inter-operate with a variety of non-military agencies (e.g., NGOs, civilian government agencies, UN organizations, and so on). Furthermore, HADR operations are, by their very nature, typically undertaken in very hostile and dynamic information environments, environments where relevant information is incomplete, uncertain and liable to change. These features clearly pose a range of information integration, exchange and interoperability challenges for military coalition forces, and in order to deal with these challenges, the AKTiveSA project explored a number of knowledge-based capabilities, each of which was built on top of the Semantic Web.
In addition to the focus on semantically-enabled capabilities, a significant proportion of the AKTiveSA project was devoted to an exploration of user interface design issues. This was reflected in the main technological outcome of the AKTiveSA project, namely the AKTiveSA Technical Demonstrator System (TDS). The AKTiveSA TDS is a Windows-based desktop application that combines a 3-D information visualization environment with a variety of graphical components to support the adaptive retrieval, manipulation and visualization of operationally-relevant information. The following narrated videos provide a brief overview of the functionality of the AKTiveSA TDS (simply click on a thumbnail image to access the video). Alternatively, some screenshots of the AKTiveSA TDS are available here.
The scientific and technological outcomes of the AKTiveSA project provided the basis for a number of subsequent research projects. These included the SEMIOTIKS project (which focused on information extraction, knowledge processing and information visualization) and the MIMEX project (which focused on issues of cognitive augmentation, trust and the sharing of information across organizational and cultural boundaries).
The AKTiveSA project was completed in December 2006. It was undertaken as a stand-alone project as part of the UK's Data and Information Fusion Defence Technology Centre (DIF DTC) initiative.