Demonstrations

 

Below is a selection of our digital services innovations.  Demonstrations can be arranged on request either at our offices or yours.  Contact Annette Dockerty via email annette.dockerty@smartservicescrc.com.au.

                                       

 

Real-Time 3D Modelling

The Real-Time 3D Modelling project aims to develop key technologies for creating 3D models of individual objects and indoor scenes using a low-cost RGB-Depth camera. Prototype software tools shall be developed upon the key technologies for users to capture 3D models with true colour texture by simply holding and moving a single RGB-Depth camera around an object or within the scene.  Currently, a preliminary version of the software tool has been developed for real-time construction of full 3D object models and coarse 3D scene models. Real-time mapping of colour texture to the 3D models and construction of refined 3D models of indoor scenes are being developed.

 

Markerless Technologies (SmartLinkIt/AirLink™)

The SmartLinkIt application enables a user to hover over an advertisement or an image accompanying printed media using an iPhone or iPad camera, and automatically trigger the delivery of additional online content, for example video relevant to the ad or the printed story. The underlying technology developed by RMIT and UoW researchers relies on open source machine vision algorithms that automatically process and detect features that represent a “fingerprint” of an image captured by the camera. AirLink™ is the first commercial application of SmartLinkIt and is currently a feature inside the SMH mobile application. This approach does not require a user to click to take a photograph and does not rely on a marker, such as a QR code.

 

EyeTracking Technologies (Focus)

The 'Focus' technology allows an app to monitor the gaze and attention of a user on an iPhone or iPad. There are various applications such as increasing size of articles upon gaze, linking to external content (as with SmartLinkIt & Airlink™), providing real-time analytics on customer viewing behaviour etc. The underlying technology developed by RMIT and UoW researchers is unique in that it avoids complex calibration requirements.  Hence, 'Focus' is usable by any user without setup requirements.

 

ServAlign

ServAlign is a tool for organisations to model their business strategies, decompose strategies into objectives and sub-objectives, and eventually map these objectives to the portfolio of services they already have in place, or the portfolio they need to create in order to meet these objectives.

 

MediaWall

MediaWall provides a novel way for the public to engage with a new form of gesture-controlled content on a large display, projected on the side of a building or any large surface. The demonstration shows how MediaWall is designed to grab people's attention out of the corner of their eye, with an engaging and innovative way for users to browse through the available media (for example, video advertisements related to the retail outlets in the building). While applicable to a range of industries, developed by the University of Sydney the MediaWall has the potential to combat the challenge of "display blindness" faced by current digital signage systems in public spaces.

 

Extraction from Multimedia

A Smart Services research team at QUT is applying its expertise in facial recognition technology to identifying and indexing video news archives. The technology combines multiple techniques to identify the personalities featured in a news video report and will allow publishers to search their historical news archive for specific people (for example, to find all videos where Julia Gillard and Kevin Rudd appear together). This data can be used to drive video recommendation services, provide editorial decision support, and support new news delivery technologies.

 

News Portal Processing

The News Processing Portal (n.pp) draws together for financial analysts and media advisers a number of web services that can be used to create, manipulate and visualise news datasets. This enables users to track the effect of changing news and trends for future forecasts of outcomes. A prototype is being developed with Sirca, a leading Australian provider of financial data and eResearch services worldwide.

 

Recommender Technology

Working with researchers at QUT, UNSW and USYD, Smart Services CRC has extensive expertise dealing with Big Data, data mining and machine learning. One particular application of this expertise has been developing recommender systems to enable one-to-one marketing and personalised service delivery. Smart Services researchers have developed recommenders for areas as diverse as personal dating, cars, movies and road surface maintenance. Many of these applications require different approaches that go beyond the well-known “collaborative filtering” technique popularised by product websites such as Amazon.

 

Life-Event Handler (iPhone app)

The Life-Event Handler shows how companies can leverage IT to create a secure channel to be closer to their customers and become more pro-active in sensing and responding to life events. The life-event handler enables customers to delegate the translation of events into actions to organisations (no more “what do I need to do now?’), and provides support for managing their personal processes and private assets. The Life-Event Handler is developed by QUT within the Business Service Management project in collaboration with the Smart Services CRC Foundry.

 

LabMET Predictive Decision Support for Health Services

Using laboratory based information already available as part of routine patient care combined with electronically held patient demographic data the LabMET service extends current validated predictive models into “Real time” patient alert system. Currently in live trials at Austin Health, Melbourne and in collaboration with researchers from the University of Melbourne, LabMET identifies patients at risk of clinical deterioration 10-24 hours prior using a modelling prediction based on laboratory blood work/data and historical data of Medical Emergency Team calls, cardiac arrest and ICU admission. These models have been developed and validated with good predictive sensitivity and specificity at 2 Melbourne hospitals.

 

A Data Analytics Application for Mining Sensor & Corporate Data

Utilising data mining expertise from QUT this project collaboration with Qld Transport and Main Roads (TMR) looks to provide decision support to identify suitability of new deflection measuring equipment operated under Australian conditions and to identify factors affecting its performance. If this equipment is proven to be suitable, data collected from this equipment would support TMR implement a more effective road maintenance strategy.