Big data as a big problem
Everyone involved with the assessment or appraisal of land knows how much information is available for sifting, processing and interpreting. Indeed, in 2016 Spring Terrier, the article “Digital Living – Understanding the Future of Cities and Public Services” raised the issue of “Big data” and the problems that skills deficits are creating in the effective use and interpretation of such data.
Old data was hard to find
However, it was not always so; while the proliferation of data today seems normal, it was not so long ago that site specific data was hard to come by.
Graduates in the 1990s often spent many hours poring over historical parish maps and county records in libraries around the country – many of whom either did not allow, or provide photocopying facilities, meaning that travelling with a pad of tracing paper and pencils to make your own record of the maps was crucial. Information from ledgers would be copied out in long hand.The answer to this rather analogue approach was of course Geographical Information Systems, or GIS, as it was widely known. Once upon a time this was rather obscure technology. It was taught in a laboratory environment in university basements and required considerable programming and coding skills. Run on a command prompt screen, development of the systems took time and effort, and this even extended to operation as an end user.
Welcome to the future – all you can eat data
Now, however, GIS is mainstream technology, although it often does not speak its name. Consider Google Maps. This allows a user to see all the key geographical features in an area, locate buildings, determine what businesses are there and thanks to the data collected by users of phones equipped with Google Android Software, can provide real time traffic and pedestrian footfall data. All of this can be presented on a fully scalable map, or if you prefer, an aerial photograph that is updated regularly. In fact, the problem now is not too little data, it is too much. The creation in 2010 of the Open Information Licence by the UK government and the subsequent launch of the website www.data.gov.uk which saw the release of all non-personal government-held data, was a game changing event in UK site analysis.Whereas once there was a scarcity of information, and a lack of portability and processing ability, there is now an exciting, yet intimidating, amount of information available. This “embarrassment of riches” leads to information overwhelm and potentially “paralysis by analysis”.
The development of GISSMo®
This was a problem that we encountered while undertaking regeneration schemes on complex sites. It became apparent that lots of geographic data was available, and with a multi-disciplinary cross team approach, data centralisation became a key consideration. We found that initial attempts to have a weekly data issue (on CD) to the team was unwieldy and time-consuming to produce, while providing no control over which data versions were in circulation and in active use at any one time. Data security was also an issue, with confidential and privileged information being issued on the discs. We therefore needed to develop a method of centralising our data while ensuring accessibility. Our target benchmarks for
the solution were:
- Central control of data
- Ease of distribution
- on-technical user focus
- Highly interpretative user inter-face (or UI, in software parlance).
Former RNAS Daedalus – remediation and site disposal project
The GIS team at CampbellReith took on this challenge and the result was GISSMo® – or Geographical Information Systems Site Model This proprietary software comprises an online GIS system that runs on the Eartlight software, developed by StatMap. This can import any geo-referenced database, such as those from magic.gov.uk, the Environment Agency, data.gov.uk, or your own in-house property data (as an example CampbellReith holds a database of the lost rivers of London –now buried by development – which we created by digitising historic maps). Individual users are granted a unique ID and password and their access to certain datasets constrained according to need. For example, technical specialists looking at ecological issues would have no need of seeing information relating to the projected land values, so their ability to do so can be prevented by the administrator.Data can also be made available to third parties outside of the project team. During land purchase negotiations, it is possible to allow a prospective buyer access to key data sets. This user cannot download or alter any data and can also be confined to data within a particular area, be that the entire site boundary, or a discrete land parcel within it. We have used GISSMo in this manner on Homes and Communities Agency (HCA) sites – indeed it was this specific need that drove the development of the system in the first place.
The utility of GISSMo is best demonstrated by looking at some examples:
RNAS Daedalus, Portsmouth
This project represented the genesis of GISSMo. Faced with a complex remediation project on a vast former Royal Naval Air Station in Lee-on-
Solent, and the need to provide parcel specific information to potential purchasers, GISSMo was created to provide a highly secure, controllable yet
flexible data sharing environment.
Partnership for Urban South Hampshire (PUSH) Spatial Strategy
A strategic regional study to inform the development of local plans for 12 local authorities in the South Hampshire region, we used GISSMo to assess development constraints and subsequently identify potential development locations for 120,000 new houses. The number of councils and officers involved meant that GISSMo provided a vital tool for collaboration.
Earls Court, London
As a slightly different example, in our long-term role as engineers to the Earls Court Exhibition Centre in west London, GISSMo was deployed to provide a repository of historical drawings, many going back several decades. Data relating to portal frames, beams and columns was collected and geo-referenced along with floor plans, underground tunnels and photographs from inspections. This made location of records possible by the simple clicking on a map. This information has been used to inform modifications and developments, and ultimately for the demolition for the site without impacting on the underground tunnels which lie beneath the Earls Court footprint.
Benefits of GISSMo®
The key benefits of GISSMo are that is removes the need for in-house, or in-team, GIS technical expertise. Anyone with a basic understanding of computers can use and benefit from the system. The approach also offers excellent scalability with the ability to authorise multiple users to the same system. The system also avoids the need for GIS software; the online system is available on any internet enabled computer and can therefore be
used from any location with an internet connection. Mobile use is already in development. GISSMo also determines a reduced time to benefit. It can be made “up and running” quickly, with even unfamiliar users being able to interrogate the system within minutes. The intuitive nature of the UI makes handling large volumes of data far easier, and reduces overwhelm when presented with new data sets. Finally, data is stored and maintained centrally, allowing you to exert ultimate control over your information.
Geographical information in the future
What is clear is that the “information glut” is not going to go away, and will only get worse. This will be caused not only by the capture of new information, but as legacy data systems are increasingly established, bringing veteran data sets into the usable realm. For example, CampbellReith is currently digitising its old drawings and specifications so that they may be available for reference by our engineers and designers of the future. Many other organisations are engaged in similar projects.
From database management to Information Curation
The next step for so-called “Big Data” is therefore Information Curation, where the focus will not be on volume, but on currency and relevance. The way in which we like to view the development of information is using the Data Hierarchy (McGawley and Ilet, 2010). This shows how data matures from raw numbers to actual wisdom, with increased interpretation and application. Increased curation moves us along the hierarchy, improving the quality and value of the data to us, the users.
Earls Court, London – Structural information stored as geo-referenced data
Empowering the user
The increase in user-focus will continue, as UIs are improved and respond to actual usage patterns, rather than those prescribed by software designers. This will probably take the form of smart UIs that adapt and update themselves to reflect usage patterns. The development of more advanced search algorithms will make cross-referencing datasets faster with more accurate results. The technology already exists for mobile GISSMo access, allowing data to be available to users on site via their smartphone or tablet, and will be incorporated into the system very soon. A further and potentially even more exciting development is the possibility for this to evolve to real time online editing of data, allowing surveyors to create new data sets on site, which can be shared with almost immediate effect with other members of the team.
(Data) failure is not an option
Perhaps the loudest message coming from these trends and changes is that data failures will not be tolerated in an increasingly data literate world. The expectation for seamless data delivery has been established and performance is expected.We are now seeing a move along the Data Hierarchy that is enabled by technology, meaning that Geographical Information Systems are becoming Geographical Knowledge Systems, delivering higher utility and value than they were ever originally designed to do – as former basement dwellers learning GIS coding, we can certainly attest to this. The development of GISSMo is a step further in this evolution, and although we are probably some way from our first Geographical Wisdom System, we will keep working on it.
Spencer McGawley & Rebecca Pembery
Article published in The Terrier – Autumn 2016
Click here to see pdf version