OGC and Grid Useful Technologies for GMES

1
OGC and Grid -Useful Technologies for GMES?

• Richard Saull,
• Adrian Grenham, Stuart D. Fowell,

2
OGC and Grid -Useful Technologies for GMES?

• What did we do and why.
• Looking at GMES from another angle
• What are GRID and OGC?
• When will they be ready for GMES?
• Technology Readiness
• Where could they be useful in GMES?
• What are the priorities for development?
• Conclusions Does it matter?
• I have assumed you have some knowledge of GMES
• Only high level points can be made in 15 minutes

3
What and Why

• Why Bother ?
• GMES funding bodies now nearing point of funding
  systems and infrastructure need to understand
  whether new technologies can enable GMES services
  and what has to happen for them to be accepted.
• Aim and approach
• Practical GMES requirements driven Focus on ESA
  GSE services
• Use Technology Readiness Level (TRL) methodology
• Outputs
• Assessment of likely usage of OGC and Grid in
  GMES architectures
• Recommendations for development order
• Why SciSys?
• GMES system integrators on 2 ESA GSE projects
• Providing lead on another project
• In touch with Environmental users
• Experience of Grid via technology studies
• Experience of GIS via implementations for
  institutional users

4
Looking at GMES from another angleSystems
Infrastructure

• What is under the bonnet?
• How do these services work
• (rear or front wheel drive?)
• What are the important bits?
• Wheels, Engine, Transmission, Wiring
• How much fuel do they need
• How much power is needed to get up the hill?
• 12 ESA GMES Service Elements
• Same statement of work
• 12 Different approaches to system definition

5
GMES Service Delivery Chain Overview How does it
work?
Data Provider Domain
Satellite
EOData
6
GMES Service Types
1 Push Regular scheduled delivery of
information
2 Pull information need triggered by
external irregular event
2a Mixed standard information triggered by
external irregular event
12 GSE Projects - 59 Services, 47 Push, 12 Pull
7
GMES total data demand (not all 10 services)
NB Overlap Cloud
8
GMES Architectural Characteristics

• Relatively static supply chain applications VO
  based
• Web Service-based applications tend to be very
  dynamic in terms of service selection, NOT the
  case with GMES
• Traditional EO supply chains tend to be very
  ad-hoc
• Each organisation may be in multiple VOs
• Access security depends on particular VO, seen as
  key
• Service level agreements within VOs statically
  agreed
• Large, geographically dispersed data archives
• Standards (metadata, cataloguing) key to
  interfaces
• Efficient data transfers from/to archive,
  optimised data caching of inputs, intermediates
  and outputs
• Computationally intensive data processing
• Distributed processing can speed up processing
  where the job can be split into many smaller
  parallel jobs
• Varied product generation and delivery
  performance requirement
• Floods to urban sprawl
• Work flow generation or selection of higher
  order data products?
• Cyclical, long life, including operator
  interactive steps

9
What are OGC and GRID? In one slide..?
10
When will they be ready for GMES? Assessing
Technology Readiness

• NASA, ESA and UK MoD use Technology Readiness
  Levels (TRL) to define / manage technology RD 9
  Levels, ranging from initial idea to fully
  robust, validated solution

11
How ready are OGC and GRID?
12
OGC GRID Where could they be useful to GMES -
Soon?
- Geographic Data Interface - OGC is applicable
13
GMES opportunity near archive Grid enabled
pre-processing
Satellite
Earth Observation Data
Process / Interpret
Terrestrial data
GMES Pre-Production
Mosaic etc
Information
Order / Pay
Process
Discover
Delivery
14
OGC for GMES Summary

• Enables GMES VOs via standardised spatial data
  access exchange
• GML now adopted
• WMS allows interfaces between different vendors /
  projection systems
• WFS gives potential for attributes and in-situ
  data to be accesses across the open service
  partnership
• Can be used as a delivery mechanism
• Fits in with ESA plans for ground segment
• Oxygen
• Some Security Concerns
• Not so important as security can be wrapped
  within VO

15
Grid Conclusions

• Q Would you trust your business on Grid?

• A Not yet!

• Until Grid moves from academic circles to
  commercially-supported COTS, it cannot be
  considered for GMES in the large
• It must demonstrate a step change in
  functionality before it will be accepted into
  operational GMES services.
• However can be considered within individual
  organisations for individual steps within product
  chain, e.g. computationally intensive data
  transforms
• But they must be prepared to support their Grid
  infrastructure themselves

16
Grid Technology lifecycle (Gartner)
Visibility
Peak of inflated Expectations
Sci/Tech Grids
Plateau of Productivity
Slope of Enlightenment
Commercial Grids
Trough of Disillusionment
TechnologyTrigger
Maturity
17
Summary Status Development Priorities
Grid practical GMES demos FP6?
Extend / Prove Reliance of openSource
Support Mechanisms? Cheaper COTS?
18
OGC and Grid - Useful technologies for GMES?
Conclusions

• GMES requires infrastructure
• Complex VO based system
• Service level agreements are basis for
  information supply
• GMES is reaching the stage where Infrastructure
  is needed
• That builds on emerging services
• That build on existing systems and strategies
• UK industry is well placed to exploit ESA and EC
  funding
• OGC is ready and mature for GMES
• Can enable the open service partnership concept
• Grid has a way to go to fulfil its potential for
  GMES
• Not mature enough for business
• Security
• Support
• Isolated areas may benefit now (pre-processing,
  distributed processing within processing chains)
• DOES IT MATTER ? Will anyone actually use it?
• Cultural Issues need close examination
• Accountants and Security men have influence in
  business
• What benefit do I get investing in it? Needs to
  be a step change
• Is it secure?

19
OGC and Grid -Useful Technologies for GMES?

• Thank you for listening