COALESCE: Deployable Peer-to-Peer Collaboration System

The COALESCE project was carried out by Prior Kain Ltd to demonstrate a potential candidate technology for peer-to-peer collaboration systems in the event of a loss of infrastructure or in challenging environments. The research project looked at leveraging large commercial developments in civilian technologies to build systems which are better value for money, more flexible, extensible and quick to deploy.

The COALESCE demonstrator was developed using open source and open standards in order to facilitate rapid future development and to leverage existing infrastructure where available. It demonstrates how such systems can improve situational awareness, location intelligence and foster collaboration within teams and between multiple agencies. The COALESCE project demonstrates improved information flow across the operational hierarchy during an operation and thereby also enhances the ground truth to support future planning and decision making. COALESCE also demonstrates how such a system could be rolled out in RF hostile and power free environments and potential fail over methods in the event of a loss of expected communication bearers.

SITEMorph

Sponsored by the Welsh Assembly Government under their SMARTCymru scheme, SITEMorph is a concept that looks at using machine learning algorithms to analyse the behaviours of internet customers for future targetted marketing. SITEMorph will look at finding both the right demographic for the sale of products or services on-line and also deliver a message tailored to the characteristics of each individual visiting the site.

SITEMorph looks beyond simply suggesting similar products, as with recommender systems provided by Amazon or Tesco, or targeting a specific demographic like age or location. SITEMorph includes individual character attributes, which affect our desire to buy, and modifies the content and message medium to suit particular character types. Such individual character attributes include, but are not limited to, psychographics, behavioural attitudes to making a purchase, emotional responses to purchases, social status and buying power. The main advantages of this type of system to the advertising industry will be that is it much easier to talk to the target audience in a language they understand and therefore easier to influence them into buying.

Prediction of Intent(JANUS)

JANUS 'makes sense' of apparently disparate sensor events in a scenario, looks backward and forward in time, and combines this information with background knowledge to classify, interpret and predict intent. JANUS uses a knowledge base to validate assertions, provide context, and understand the latent 'intent', and notifies the right person at the right time throught the right medium. The knowledge base can include anthropologically derived notes on cultural behaviour, models of social networks, and known interaction types. Overtime this knowledge is likely to improve along with improvements in sensor feeds so JANUS will be knowledge, scenario and sensor agnostic. During operations, patterns that are found and verified can be automatically or manually fed back into the knowledge base to keep the information current. In principle, the JANUS system can also be extended to more than one platform with automated knowledge pooling across platforms.

'Digital Conference Space' (DCS)

DCS is a social networking style service for conference or enterprise communities that augments physical world discussions with an ongoing series of digital conversations and interactions between members. We offer the secure exclusive or inclusive DCS tailored to your requirements and branded and styled in line with your branding. DCS functionality includes: discussion forums, photo galleries, links to other social networks (e.g. Flickr, YouTube etc. ), browse and search facility, geospatial capability, tagging and tag clouds, compliance with XHTML-Basic mobile devices, on-line advertising, organisation profiles, public and private spaces, basic content management (CMS), user profile pages, accounts and roles, administrators/moderators pages, user-themable blogging, central news’ blogs with RSS feeds, email functionality, notifications and opt-in newsletter functions, 'Facebook Wall' style broadcast mechanism, and a double-blind friendship system.