City Magazine 2014
  1. Dispatches
  2. Discoveries
  3. Community
  4. Features
  1. Painting the bigger picture
City Magazine 2014

Painting the bigger picture

The giCentre at City has worked with partners including Transport for London, the Ministry of Defence and Nokia to help them make sense of their data through interactive, creative and elegant information visualisations. The giCentre recently turned its attention to energy, allowing one leading supplier to prepare for an approaching data revolution in the sector.

Glyphs for exploring crowd-sourced subjective survey classification from Jo Wood on Vimeo.

The old adage maintains that "a picture paints a thousand words", but for Professor of Visualisation Jason Dykes, part of the challenge for the emerging discipline of information visualisation is showing just how powerful an image can be. "When we are trying to make sense of data, we are more accustomed to dealing with numbers. However, a visual representation of data can often be much more successful than a set of numbers in helping us recognise patterns, place data into a wider context of time and space and solve problems." In a world where humanity now produces as much raw data in 48 hours as it managed to accumulate in its entire history to 2003, visualisation represents a powerful tool for governments, companies and individuals alike.

A collaborative approach

giCentreA recent partnership with E.ON, one of the UK's leading energy providers, provides an example of how the giCentre's techniques help businesses establish what they can do with their data. The UK Government has committed to installing smart meters, which give real-time feedback on energy usage to users and suppliers, in every home by 2020. For suppliers, the real-time readings generated by smart meter technology will provide more data than ever before about energy consumption. The giCentre worked with E.ON's energy analysts to develop four visualisation prototypes (using sample data from a group of 100 homes) that would show peaks and troughs in daily demand, compare modelled to optimised solutions (to demonstrate, for example, the impact of energy-saving technologies), group consumption by appliance and identify trends and anomalies. The visualisations offered E.ON a snapshot of how it might harness the power of its data in the smart home era.

Collaboration was also at the heart of the relationship between E.ON and the giCentre. "With all our end-users, the building of relationships and the establishment of trust is crucial," explains Dykes. "In our requirements workshop, the City academics worked in partnership with the E.ON energy analysts to establish, through a variety of creative techniques, how visualisation could help the analysts make use of what initially seemed to them to be unmanageable data." For the giCentre, one measure of success is when end-users begin to see how visualisations can help them answer existing questions using their data. But a more important sign of a "job well done" is when visualisations lead partners to ask more sophisticated questions and begin to re-shape business practice by seeking answers to these.

Global reach

As befits a group of academics whose backgrounds are predominantly in the field of geography, the giCentre is embedded within a global network of data visualisation experts. The E.ON project was undertaken in partnership with the IMDEA Energy Institute in Spain and the giCentre regularly collaborates with other leading data visualisation groups in countries including France, Germany, Japan, the United States and Australia. In March 2014, the giCentre was awarded €500,000 of EU funding to explore the potential for the use of visualisation in criminal intelligence analysis, part of a wider €13 million VALCRI project involving 18 other European partners.

That the VALCRI project - which will involve mapping data of different types and from different sources to help intelligence analysts develop hypotheses and establish narratives - sits alongside the E.ON project in the giCentre stable is an indication of the applicability of visualisation techniques across the breadth of human activity. Some of the giCentre's best-known work uses Transport for London's data on the Cycle Hire Scheme to map travel patterns in the capital, while other projects have explored voter bias in London local elections and involved working with the Ministry of Defence to help the military understand local populations (Human Terrain Analysis) through visualisations. Indeed, the giCentre's work is by no means limited to human activity: recent projects, led by Dr Aidan Slingsby, involved collaboration with animal ecologists to design and implement visualisation techniques to understand the daily movements of seabirds and the migratory characteristics of geese as these birds move through time and space.

For Dykes, who will chair IEEE Information Visualisation, the preeminent conference for visualisation researchers and practitioners at the end of 2014, some of the "big questions" dominating the discipline at present relate to how visualisation can reflect uncertainty: both the measured and immeasurable unknowns present in data sets that can be so key to our interpretation and use of data. There is also increasing interest in personal visualisation, with the advent of smart watches and other trackers that can monitor our lives, from our sleep to our blood pressure, heart rate and physical activity, presenting new challenges in how we convey ever greater volumes of data in our daily activity.