European GI
R&D Meeting
20 June 1996
Meeting Report (Final)
Table of contents
Introduction
Meeting Structure
Input from Participants
Meeting Conclusion
Background documents used during the meeting
GI2000: Towards a European Policy Framework for Geographic
Information (Key Issues)
Recommendations for the 5th Framework Programme
Potential Fields of GI Research in the EU 5th Framework
Programme
GI/GIS Technologies
GI Applications
The GI (Geographic Information) R&D experts meeting was convened by
DG XIII/E.3 on 20 June 1996 as part
of the wider consultation process regarding the development of the document
GI2000: Towards a European Policy Framework for
Geographic Information and in order to elicit further comments from the
research community in regard to GI content for the Fifth Framework Programme
development exercise. Annex A contains the draft agenda. The list of individuals
and organisations invited is contained in Annex B, which also shows those
who actually attended.
The latest version of the GI2000 Policy Document,
dated 31 December 1995, was distributed at the meeting, as amended by M
Littlejohn of DG XIII/E and R Longhorn, external expert in GI, specifically
to take account of comments from the previous meeting with the Commercial
GI Data Providers (held 28 November 1995).
The meeting was opened and chaired by Mr. Martin Littlejohn, DG XIII/E.3,
who also used this opportunity to provide a progress report on the INFO2000
Programme, which was approved by Council on 20 May 1996 and for which the
first Call for Proposals was expected to be published the next day (21 June
1996). It is under INFO2000 that certain
specific actions relating to further development of GI content in Europe
will take place.
R Longhorn also distributed a short questionnaire
to the 27 participants (21 of which were returned) asking for their opinions
on 5th Framework Programme content in regard to GI. The questionnaire and
brief analysis of the results are contained in Annex D.
Slides used by Mr. Littlejohn and Mr. Longhorn in their presentations of
GI2000, INFO2000 and the Framework Programme development are in Annex E.
Mr Littlejohn (ML) called the meeting to order at 0930h, welcomed the
participants, and outlined the day, which would focus on GI2000 in the morning
session and on the Framework programmes in the afternoon. He also provided
an overview of the INFO2000 programme, preparatory work for the 5th Framework
Programme, and presented the main objectives of the meeting in regard to
GI2000. R Longhorn presented proposed "key issues" of potential
concern to the European GI R&D community with regard to GI2000, which
were offered for further discussion by the group. (See Annex C).
A tour de table followed, focusing (as much as possible) on GI2000 issues.
After the lunch break, Mr. Littlejohn made a short introduction to the draft
input to the 5th Framework programme in regard to GI, which DG XIII/E.3 has
made to date, and explained the procedure and timescales involved in creating
a Framework programme. Mr. Longhorn then presented more detailed slides on
the 4th Framework Programme and on proposals (so far) in regard to setting
the agenda for the 5th Framework Programme. Following these two presentations
(key issues in Annex C and slides in Annex E), general discussion was requested
from the floor. Many issues had already been raised in the morning session.
Participants made interventions both during the morning in the tour de table
and in the afternoon by direct intervention. During the tour de table,
participants were requested to make short statements regarding their key
concerns as GI researchers and/or as product developers using GI/GIS.
Rather than report what each individual said, this report contains a summary
of issues raised or main comments made on the principal themes. These are
listed in no specific priority order, nor is any attempt made to separate
comments and/or issues as pertaining to the
"GI2000" policy document or to
the "5th Framework Programme", as
the two were often mixed.
- Price of GI must be lower to help market grow and for those (non-commercial
users) who need access, e.g. for social purposes such as environmentalists,
but who cannot afford to pay.
- R&D should be market driven and applications oriented in order to apply
limited European R&D funds for maximum benefit to the overall GI market.
There are many practical needs already apparent from the discussion which
could be resolved in the near future by a concerted effort in specific programmes
of research.
- If GI R&D is more market led/industry focused, then it may be more acceptable
to politicians who see nearer term return for their research funding than
in the past. It also forces industry and research communities to focus on
real problems of importance to the market. (This approach is taken today
in the USA via the current approach of the National Science Foundation working
more closely with industry).
- European GI base data needs to be defined now! "We have been discussing
this for two to three years now, with no results. Time to simply get on with
it - and there are models already in existence, such as US FGDC "framework
data", which will provide valuable clues as to how to proceed."
- Need to put more emphasis on the data updating problems. There should be
better mechanisms in the GIS and database toolsets and there should be more
research into ways to lessen the cost of updating existing GI datasets. For
example, it should be possible to integrate new data into existing datasets
without being required to make a wholesale replacement exercise, often of
data that has now had value added to it since it was originally entered into
the digital database or last updated.
- GI data quality is still in its infancy. Need to educate the user (especially
new users) as to quality issues, which are many. Need practical ways to establish
data quality and to transmit objective assessments of quality to users. This
is a primary topic. Need to establish a metric for measuring and reporting
quality. There should also be standards developed for measuring quality.
Must examine what users want in regard to "quality", not just what
the data holders are willing or able to provide.
- Data needs to be more readily available and less expensive. "Content"
also needs to be better defined, as per real user needs, not simply "what
we have available" as a result of previous data collection efforts.
- Remote sensing data - need to find automatic tools and/or new techniques
and methodologies to help exploit the wealth of raw remote sensing data that
already exists, let alone the volumes of data that will be produced by newer,
higher resolution satellites.
- Cadastral data is a primary focus, especially at local government level,
and too little is being done with GIS in this subject area in many countries.
- Techniques and technologies to permit generalisation and/or creation and
use of scale-less databases still needs much R&D. The benefits need to
be better defined and the real "markets" for end-products identified.
The problems will not be resolved soon!
- The spatial component of the "all pervasive" GI is often difficult
to capture and/or to maintain.
- You (the user) can already buy GI directly from Internet or Web servers.
The ramifications of this need to be more fully and formally investigated,
especially for naive or first-time users who lack adequate foundation training
in many of the GI data related issues listed above.
- Do we focus too much only on "data"? Toolsets available now make
it easier to manipulate data, yet training is needed to properly interpret
the results. Data provides possible "answers" to questions, yet
the questions come from concerned users, many of whom do not understand GI
or spatial referencing or the power of GI tools to help them focus on their
problems.
- Standards can be interesting when introducing GI to a "green field"
site, i.e. a completely new user. Several standards are still missing from
the field of GI (no further elaboration as to which).
- Interchange of data is still a major issue - expensive, time consuming
translations and transformations. To integrate data in Germany from the 16
Lander required creating 13 different conversion programmes. Yes, it can
be done, but what a waste of time! The EU should strongly support the development
of public domain data formats and data exchange procedures.
- Interoperability of data and software must be enhanced, across multiple databases
from multiple sites in multiple formats and even across multiple hardware
operating systems - and, in very near future, across the Web.
- Integration issues need more emphasis for GI than for other forms of multimedia
content, because of the nature of GI.
- Metadata is needed - now, in format and with breadth of coverage suited to
the needs of the market. Provision and use of metadata needs a more rigorous
test bed than is apparent today, from organisations simply putting up random
metadata collections without regard to standards, needs of users at different
levels of proficiency, etc.
- CEO (Centre for Earth Observation) should be monitored and perhaps used as
a model of how to implement GI metadata and consultancy services in areas
outside "earth observation".
- What do users really want and need? No one has properly researched this yet,
either for GI itself or for its metadata.
- More and better public relations (PR) is needed in order to get the political
support needed, not only at European Commission level, but at national level
as well. Need to stress and demonstrate how use of GI and modern GIS tools
can help solve difficult problems (networks, planning, environmental control,
etc.) - and this message needs to get to the decision makers, most of whom
are not "GI aware."
- Hand in hand with PR goes more training, at a range of levels, from short
courses for real "experts" to introductory courses for complete
beginners, along the lines of half-day introductions to word processing and
spreadsheet programmes.
- The GI data requirement is four dimensional, not two dimensional (which is
the limit of many current GIS tool sets) or three dimensional (which is the
limit of the latest GI technology). Temporal aspects of GI need more focus,
and more temporal based features need to be added to GIS toolsets now. Many
problems can only be examined when GI is examined over time, not just across
living space.
- Visualization of maps and content (both simple and complex/complete) are
key issues with groups such as motoring clubs and tourist organisations.
People who buy such products want the "digital map" product to
"look like paper maps".
- More work is needed in integrating multimedia and GI/GIS, especially latest
techniques such as virtual reality. Also, as multimedia moves more widely
onto the Internet, via Java and HotJava on the Web, GI data holders and users
need to keep pace with these developments.
- Applying advanced multimedia concepts to GI presentation can also help increase
awareness of the power and utility of GI.
- Need better and simpler GIS tool sets with more extensions. It is a dangerous
fallacy to think that "what we have today in GIS toolsets is all that
we need for some time to come." Current tools are OK for basic map
functions. However, as soon as that "next step" is needed in the
analysis process, e.g. cross-border analysis using different national datasets,
then the user normally must revert to purpose-written models, spreadsheets,
etc. to achieve real analysis capability.
- The GI researchers need to follow more closely advances being made in digital
library research, generally, as the issues faced and techniques and technologies
being developed to resolve large problems, have direct parallels in the GI
world.
- Time and space modeling needs to be expanded and made easier, i.e. more R&D
on simulation software which integrates GI easily.
- Multi-scale modeling needs much more work and testing in practical applications.
- "Models" and "data" are interlinked and more note needs
to be taken of the interrelationships, especially when a model is developed
for/from one type of data and then attempts are made to use that model with
different data sets.
- Need to find ways to combine multiple models, often developed by different
researchers for different purposes, in order to tackle new problems at least
cost. Yet the diversity of the many application areas in which "all
pervasive" GI can be found makes this goal very difficult. Is some form
of standardisation possible or the answer?
- Awareness of copyright and IPR issues are important but less so to R&D
community. Collecting and maintaining GI can be a very expensive business.
You cannot afford to let it be "stolen" any more than you can afford
to "give it away" - so practical protection techniques are mandatory
before wide dissemination of GI will occur digitally.
- Must guard against misuse of data. The confidentiality issue is important
to the ordinary citizen and GI can be used to "pinpoint" individuals
or groups more so than many other kinds of data (or technologies). If there
were technological means available of protecting against potential misuse,
then GI might gain wider popularity in certain application areas.
- Creating a European network of other research institutes should be considered.
- The "remote sensing community" is part of the wider GI community,
but this does not always seem to be acknowledged or capitalised upon. Better
and/or more formal coordination is in order. Many geographers/cartographers
see remote sensing as an alternative to "maps", but without the
required resolution, when in fact remote sensing is an entirely different
way of looking at the earth.
- Industrial partners are important and a forum is needed to help bring the
GI "industrial" community closer to the "academic" or
research community, in Europe and globally.
- Public/private synergy is an ideal goal - but will be very difficult to achieve
in practice, since many public agencies holding GI have no mandate at government
level to release such data, in a controlled manner (which costs money!).
Private data providers normally collect only that data which is of immediate
use to themselves or their market niches, because of shorter economic timescales
(this year's profit/loss statement!).
- Being able to resolve "cross border" issues using GI is as much
"institutional" as it is technological or data oriented. Much more
awareness is needed at government level in order to bring this point home
to the decision makers.
- Data is a resource. GI is a primary data resource for Europe and is truly
"European" (as opposed to many of the tool sets and technologies
used to manipulate that data). This resource must be more widely shared and
used for the benefit of Europe at many different levels, both public and
private.
- There is no time left to "go slow" in regard to many GI issues,
i.e. defining and collecting GI "base data" across Europe, resolving
data integration issues, adopting stronger political support for interoperability
initiatives, etc.
- Data pricing and well-defined, market oriented access terms are key issues,
but cannot be easily resolved across the many different "selling"
regimes that exist in the European Union within public agencies plus the
commercial needs of the private data providers to help ensure that they still
exist five or ten years from now.
- It may not be possible to resolve the more politically or legally oriented
issues in the near term, yet we must find a way forward to expand use of
GI and monitor the political and legal "solutions" as they are
developed and adopted for other areas of the multimedia content industry(ies).
- To raise awareness more widely and more generally, there is a need to create
advanced "demonstrators" of the power and usefulness of GI, using
latest multimedia tools for eye catching presentation, focus on specific
applications which will appeal to wide groups of potential end users. The
money invested in this could also lead to developing "best practice"
guides for other developers.
- We (the "EU GI community") need to more forcefully bring home the
message that "GI aids delivery of services" - across a wide range
of services, many of which are vital to ordinary citizens (emergency services,
city and regional planning, environmental monitoring, network planning and
operation, etc.). This will help awaken the politicians at many different
levels to the need to understand the issues surrounding GI and its collection
and provision in Europe.
- The R&D needed for GI cannot wait until the start of the 5th Framework
Programme (first contracts for which would be let in 1999). If Europe has
not resolved many of the current technology-based problems in the next two
to three years, then Microsoft, Oracle or other large (non-European) tools
vendors will have done so.
- Long-term R&D strategic goals must not be lost in the rush to overcome
shorter-term, tactical goals which are industry directed. How to balance
these two sets of research objectives is not yet clear.
- For input to the 5th Framework Programme, we should not lose sight of integration
and applications oriented issues, since the current (5) topics presented
by M. Littlejohn seemed to focus only on technology.
- Loss of the word "Infrastructure" from the policy document is not
helpful (an opinion widely shared by the entire audience). Infrastructure
denotes or implies basic concepts such as "service",
"standards", "connection" and satisfying user needs -
all of which are key elements in the GI policy framework we are trying to
engender.
Mr Littlejohn concluded the meeting at 1700h following a brief summary of
the above key issues and conclusions. Participants were also informed that:
- A further revision to the GI2000 Policy Document might be expected, based
on he input from the last three consultation meetings
(GI-USERs on 28 May 1996,
GI-VENDORs on 29 May 1996, and today's meeting)
and this would be made available via the I'M EUROPE World Wide Web site
as soon as it was ready.
- It is planned that the GI2000 document will become a Communication to Council
and the Parliament covering the main issues set forth in the current document,
as revised following comments from the last three consultation meetings and
following internal consultation with other Directorates of the European
Commission.
- The INFO2000 programme first Call for
Proposals would be announced the next day and participants were encouraged
to make proposals, especially in regard to GI under Action Line 3.1.
Background documents used during the meeting
Key Issues for the R&D Community
Data availability - European GI base data and metadata
- Can new technology(ies) help collect base data more efficiently and at less
cost, so that it can be made more easily available?
- Can new technology help to more widely disseminate GI?
- How to reduce the cost to collect and maintain GI?
- Does collection, interconnection and dissemination of GI metadata require
new R&D - or will existing technology do the job?
- Is more R&D still needed in regard to standards issues - collection,
storage, transmission, integration?
- Can further R&D help us to better understand GI data quality - its
identification, its specification, ways to record quality information ,etc.
Data exchange standards and interoperability of data and systems.
- Where are we today in regard to GI standards?
- What other standards are involved?
- Is lack of adequate standards preventing more widespread use of GI?
- In interoperability achievable in next five years? (OGC/OGIS)
- Do standards help or hinder R&D? Does R&D speed up or slow down
standardisation?
Legal issues - where does "GI" depart from "multimedia"
?
- Should the R&D community become involved in these issues?
- IPR and copyright - can technology lead the way to resolving current fears
over protecting digital data?
- Liability, protection of databases, protection of personal privacy in regard
to GI - is there a role for technology driven solutions?
Training and Awareness - from research staff to the general public.
- Training of researchers - is the need understood and being met?
- How does lack of awareness by government and general public affect R&D?
- How can researchers increase awareness for government and general public?
DG XIII/E - Information Content
Subject: Geographic Information
Geographic information (GI) is created throughout the EU in various forms,
for many different disciplines and numerous applications. However, most GI
is local, regional or national in scope and does not lend itself to use and
analysis across national boundaries. There are great difficulties in using
GI data in disciplines other than that for which it was originally collected
or created. Wide spread, useful, understandable information about the existence,
location and cost of GI data is lacking. Thousands of new European GI data
products and services could be created if current problems facing potential
users and suppliers are overcome.
The strategic importance of this area has been substantiated by letters from
four Member State ministers to the Commission suggesting that the EU provide
a stronger political impetus to geographic information at the EU level, to
further the creation of seamless, homogenous digital maps of Europe for the
charting of the European territory as an important part of the development
of the EU. It is estimated that the economic activity surrounding collection
and use of European GI amounts to 10 to 12 Billion ECU, across existing GI
user and supplier sectors.
Broadly defined, Geographic Information (GI) is all information that can
be related to a location on earth, i.e. which has a spatial component. The
traditional conception of GI as "maps" is still valid today and
represents a significant sector in what can be defined as the GI
"industry". However, the broader definition of GI also includes
any data which can be, for example, post coded, or related to the centreline
of a highway, or to a specific section of a river, forest or even weather
system. Such data is created and used by a large variety of public and private
organisations for many different applications, such as geomarketing,
environmental protection, route planning, fleet management, facility management,
socio-economic research, emergency services, statistical analysis, land
management and urban planning.
While much GI can be presented to users in multimedia format, the inherent
difference between GI and non-GI is the spatial component. The image of a
Renoir displayed in a certain museum in a specific city and country has no
spatial element in its own right, yet the information about the location
of that Renoir is of great importance to art lovers, art researchers, tourist
and information office, insurance companies, and many other potential information
users.
The GI Constituency (Users, Suppliers, Integrators,
Public, Private).
Actions are needed to create wider use of GI, partly through technology and
partly through grater awareness of the existence of GI, and not necessarily
by wider collection of GI. The actions must be targeted directly at current
holders of GI and potential new users of GI. Unlike other multimedia, which
tends to have very distinct provider/producer/supplier versus user populations,
GI is often both collected or created and used by the same organisation,
for example large utility companies. However, there are also traditional
producers and users in the GI industry. The future needs of both types of
supplier/user must be accommodated in the new RTD programme.
Organisations involved in collection and use of GI include official mapping
agencies, statistical institutes, digital data providers (remote sensing,
travel and tourism organisations), universities, government at all levels
(from local authorities to national ministries for environment, transport,
agriculture, etc.), utilities (especially any involving networks of any sort),
retail chains and manufacturing industry, to name but a few. Existing market
studies divide the world into more than 25 major user or supplier sectors
for GI and GIS (geographic information systems) technologies.
Research and technological development is needed in the GI content sector
in the following major areas:
- GI data generalisation
Basic R&D is needed into converting GI from often highly specific, highly
accurate data into less specific, less accurate data. In other words, readily
available technology must be developed which permits high resolution data
collected for one application, such as land management, to be used for other
applications, at lower resolution, such as tourism, fleet management,
environmental monitoring or even educational products. Scale-less digital
data must be better understood in order to achieve generalisation.
- GI data visualisation
R&D is urgently needed into developing and applying new, high resolution,
low cost, safe visualization technologies to the presentation needs of existing
and new classes of GI user. Examples include current investigations into
virtual reality as a meaningful interface between large spatially referenced
GI data sets and naive users.
- Geospatial image & text integration
GI becomes especially useful via the presentation or visualisation possibilities
it provides to naive users. GI inherently requires the integration of textual
information (attributes of data) with images ("maps", regardless
of resolution). Large volumes of data and impressive digital file sizes (normally
megabytes) are involved in analysis and presentation. Indexing is not yet
fully understood for such integrated data sets. Data quality issues, accuracy,
precision and many other issues become important. Further problems arising
from the combined effect of these issues has barely been touched upon in
the information industry.
- GI in model building
New tools, algorithms and methodologies must be created which promote greater
and better use of GI to answer questions of importance to the community,
to governments and to European institutions. New citizens' applications
relating to quality of life, environment, local planning, interactive democracy
will be possible. The temporal aspect of GI must be better researched and
incorporated into models more widely.
Currently, main drivers for the GI marketplace include:
- rapidly improving price/performance ratios for low cost, desktop computing
equipment;
- greater availability of GI analysis and presentation functionality in low
cost GIS software (MapInfo, ArcView1, GRASS), or even in standard office
applications, such as spreadsheets (Excel and Lotus), databases (Oracle,
Access);
- emerging awareness of the potential of GI to present complex views to naive
users in many disciplines;
- awakening (although slowly) that GI has special value, often beyond that
which justified its original collection and use.
Barriers include:
- lack of adequate tools to manipulate, integrate and present the geospatial
component of data;
- lack of understanding by data owners/suppliers of how to best market and
re-use their GI, especially for uses other than that for which it was originally
collected, specifically regarding the value of, and management of, digital
IPR (intellectual property rights), distribution channels and technologies,
etc.;
- lack of coherent standards, widely applied, in collecting, storing and
disseminating GI data, which makes data integration for new products very
difficult, expensive and higher risk than for many other forms of multimedia
(new CEN draft standards need support for testing, certification procedures,
etc.);
- lack of progress on the Open GIS (OGIS) concept, which would reduce the power
of monopolistic players in various sectors of the GI industry, and the reduced
role of European GI firms in OGIS.
The RTD actions outlined above will help to:
- Improve competitiveness of the European GI industry (developers, data producers,
system integrators).
- Generate new employment potential in both the GI data and GIS tools supplier
sectors.
- Strengthen cross disciplinary synergy and public/private synergy at European
scale.
- Improve citizens' participation in direct democracy through interactive
participation and simulation in new development projects (e.g. new airports,
motorways, school location, city renovation, etc.)
The European GI industry, across many separate, identifiable sectors, is
developing without a coherent framework. Data collection is being duplicated,
leading to wasted resources and reduced competitiveness against global players
in the GI arena. Actions towards creating a consistent framework for development
of GI are dispersed, via ad-hoc initiatives at national level. Organisations
such as EUROGI, MEGRIN/CERCO, GISPE, GIVE, AM/FM Europe and others have been
created with pan-European mandates from their members, but are only slowly
beginning to work together, outside their specific sector-based foundations.
National GI initiatives are underway in the GI industry. Better coordination
is needed across both national boundaries and across industry sectors (for
both users and suppliers of GI and GIS tools). EU funding could help to achieve
such coordination, focusing such funds on cross-border pan-European problem
needs.
- GI data generalisation
- GI data visualisation
- Geospatial image & text integration
- GI in model building
- GI data capture, handling and management, including image/text integration.
- GI data integration problems, especially from different sources, different
formats, different levels of precision, etc.
- Virtual reality visualisation techniques for communicating spatial information.
- Using GI interactively, on-line ("GI on the Web").
- Temporal GI (time dependent elements) - how to capture, record, store, process
and present the temporal aspect of GI.
- Creating and manipulating scale-less databases.
- Image processing (raster to vector, remote sensing image enhancement tools,
etc.)
- Error propagation (how to determine when it is happening and how to avoid
it!)
- Other evolving data exploitation technologies, tools, algorithms and
methodologies.
- Metadata (directories) and information locators.
- Defining and investigating the use of GI data quality information (recording,
presentation to users).
- GI data content definitions - refine "base data" concept via research
and applications.
- Creating seamless databases (across national boundaries and across disciplines).
- Create new citizens' applications relating to quality of life, environment,
local planning (best use of living space), interactive democracy.
- Developing and exploiting Open GIS - interoperability.
- Emerging standards at European and global level - Validation and best practice.
- Exploiting proposed GI metadata standards (CEN TC 287) - Validation, use
and best practice.
|