Andrzej M.J. Skulimowski AGH University of Science and Technology and

1. MulticriteriaSelection of Sites for the EuropeanProjectsSupportingLocal Information Society Andrzej M.J. Skulimowski AGH University of Science and Technology and International Progress & Bussiness Foundation ams@agh.edu.pl

2. Introduction 1. Background information on the earlier projects: 1.1. The 6th FP Rural Wings project has been the selection and installation in Poland of 10the pilot sites of bi-directional satellite internet in DVB-RCS technology (carried out in 15 further countries) The Progress & Business Foundation, Krakow, Poland has been the National Coordinator for Poland responsible for the above tasks. Timeframe: January 2006 – March 2010, Selection: 2006 Installation and test runs: 2007-2010 (2 Phases: Phase 1 since March, 2007; Phase 2 since March 2008) 1.2. Interreg IIIB CADSES TELEACCESS – project aiming at selection and operation of innovative rural telecenters (Poland and 5 European countries) Timeframe: June 2006 – June 2008, Selection: 2007 Implementation: 2007 - present

3. Introduction (ctd.) 2. The motivation and the mainissue of thispaper How to assess the quality and appriopriateness of the pilot sites’ candidates, takingintoaccount the outputs of earlierprojects? 3. The approachused to solve the problem - Compare the initialselection-stage (ex-ante) and performance-based (ex-post) pilot sites’ rankings, - Calculate the distancebetweentheserankings, whichevaluates the quality of initialselectionfit, - Assessseparately the appriopriateness of criteria, the pre-selcetion of target groups, the adequacy of the outrankingprocedure, the quality of data gathered

4. The pilot sites’ selectioncriteria • The main three groups of initialcriteria : • Geographical situation of the site – according to the Project’s goals, the more remote and isolated the site, the more eligible for the Project, but the diversification of geographical locations and an even distribution of sites over the target areas should be taken into consideration as well. • Existing internet infrastructure – the pilot stations should be installed in areas without broadband facilities or – at least - the internet access provided by third parties should have lower transmission rates or be less reliable than the satellite one. On the contrary, other existing IT infrastructure, such as school computer labs, LAN etc. would be of advantage for the project. • Availability of potential end-users willing and capable to engage in the activities of the pilot station.

5. The pilot sites’ selectionprocedure • Needs assessmentin order to choose most appropriate target groups and areas. Itresultedin a pre-selectionofsites and target groups, namely: • A) Thegeographical situation of sites was restricted to thoseregions in Poland, where the problems with low broadband access was related to the topography and geographical isolation, i.e. to theBeskidy mountains in Southern Poland, thenmountaincounties intheMałopolskie and PodkarpackieVoivodships,fulfilling initial low-broadband-access assumptionswerespecified • B) The target groupswererestricted to: • - rural schools and public libraries, • - remote research stations and national parks, • - public access points in local government offices, • - localcultural centres and telecenters, • - touristic establishments in themountains.

6. The pilot sites’ selectionprocedure(ctd.) Other categories of sites considered: - policlinics - fire brigade centers in remote rural areas, (there existed already governmental plans to endow them with broadbandinfrastructure) (ii) Data gathering using the questionnaires adjusted to the specific Poland’s needs discovered in Step (i). The questionnaires were mailed to the selected target groups (schools, national parks, local authorities, research establishments, business support organizations; since March 2006). It was supported by numerous personal and phone contacts and discussions. (iii) The data gathered were verified during field visits and otherwise.

7. The pilot sites’ selectionprocedure(ctd.) (iv) Acomputationalmethod was elaborated to transform the questionnaire data into the measures of subcriteria.(v) Amulticriteria outranking method based on the reference sets approach was usedto establish a preliminary ranking of sites: A) For eachclassof potentialsites, i.e.: - rural schools and public libraries, - remote research stations and national parks, - public access points in local government offices, - localcultural centres and telecenters, - touristic establishments in themountains, there was defined an ideal (maximallydesirablefromthe point of view of theproject’sgoals) site model, represented by thedesirablevalues of criteria. B) Thesiteswerecompared to theidealmodelsinthecorrespondingclasses, thosewiththeleastvalues of the (distance-based) regretfunctionwereconsideered as potentialcandidates

8. The pilot sites’ selectionprocedure(ctd.) (vi)Finalarrangementsand ranking The selected sites, in the order resultingfromthe aboveoutranking procedure, were visited, the Rural Wings project presented to the appropriate local authorities and potential sponsors of sites. The technical feasibility of installing the DVB-RCS equipment at the selected sites, the results of negotiations, and the final assessment of the viability and usability of site derived from the field visits was takenintoaccount by the Selection Committee to establish the finalranking Then it followed (vii) The installation of satelliteterminalsaccording to the finallyderived priorityorder

9. Results: initial ranking of candidates to becomeRuralWings pilot sites in Poland

10. Phase I pilot sites’ selection – untilJune 2006 Implementation of the initialselection ranking: The Arctowski Research Station, ranked 1, could not get the satellite connection because of the technical infeasibility of the Atlantic Bird 1 B satellite to reach the St. George Island. Heavy snowfalls, that came as early as in September, made the installation of the D-Star terminal in Hornsund, Svalbard(ranked 3) impossible. Finally sites ranked 2 and 4: Babiogórski National Forest and Polana Primary School turned out to be successful winners of the phase I selection process. The otherthreesites (ranked 5-7) wereretained for a laterimplementation, but none of them was actuallyinstalled A similar procedure to the above was performed for the 2nd phase pilot sites, yielding a list of eight pilot stations.

11. Polishsatelliteinternet pilot sitesselectedwithin the RuralWingsproject

12. Polishsatelliteinternet pilot sitesselectedwithin the RuralWingsproject

13. Polishsatelliteinternet pilot sitesselectedwithin the RuralWingsproject

14. Polishsatelliteinternet pilot sitesselectedwithin the RuralWingsproject

15. Implementationof learning scenarios in pilot sites Explanation of symbols used : XP - eXtensive use during the pilot phase highly Probable P - Potential or Planned use at a later date D -to be Determined, depending on a possible expansion of the target user group N –No use expected during the project’s duration R -Research IAP-Internet Access Point T-EL -On site Teaching and E-Learning access

16. Criteria of evaluation of the Project’sresults

17. Criteria of evaluation of the Project’sresults in Poland (ctd.) Remarks: 1. The overall technical assessment criterion (col. 7 above) was a result of rounding and normalizing to the scale [0,…10] of the linear combination of the absolute transfer rates (positive weights) with the unreliability of the link expressed by the standard deviation of the down- and uplink rates (with negative weights). 2. The highest mean values received for downlink in the site POL05 were accompanied by highest connection risk that reduced the technical score. 3. The criteria 7 and 8 were, in fact, correlated, since the main reason for a lower-than-expected use of dedicated applications were the lower than expected transmission rates, that made the use of applications via satellite more difficult

18. Thefinal ranking of sites for theSoRuraLLprojectin Poland Abbreviations: TECH - Technical performance assessment, scale from 0 (lowest) to 10 LEARN - Intensity of implementing learning scenarios, scale [0…10] IS - Experience in Interreg, LLP, or FP IS programmes, scale [0…5] PROG- Progress achieved during the pilot phase, scale [0…5]  - Total score (sum of scores from cols 3-6)

19. Therecommendations and conclusions • The installation and operation of the satellite internet access points in the DVB-RCS technology does overlap with the middle-term IS policy goals, as specified in the related policy documents, and with the statutory duties of dominating telecoms, • The satellite internet can provide safe, affordable e-infrastructure with geographically unlimited access, • The successful operation of the commonly accessible and school installations depends on the availability of comprehensible and common e-government, e-learning content and e-commerce services, • The growing role of e-health-related applications, that have to be made available to everyone, without geographical limits, is to be taken into account as well,

20. Therecommendations and conclusions • The satellite internet access can foster an extension of common intellectual sphere in the e-space and eliminate the digital divide, • The permanent surveillance of access points in rural areas and schools is necessary to monitor negative aspects of IS development, such as personal alienation through e-gaming and –gambling, dependence on internet etc., that can endanger schoolchildren and internet newcomers. • The development of the IS in rural areas based on the satellite internet will support a move from the traditional agriculture to the development of innovative, sustainable, environmentally and animals-friendly agriculture that supplies high-quality end products . • Otherbranches of ruraleconomy, e.g.(agro)tourism, canbenefit from the satellite broadband access.

21. Conclusionsregardingthemulticriteriaselectionmethod • The ex-post evaluation of the initial ranking, based on the results of the pilot sites selection compared to the ex-post assessment criteria may allow to derive a learning scheme, which can be useful to assess the adequacy of the selection criteria, the adequacy of choice and correctness of using of the multicriteria outranking procedure applied in the selection process, and the credibility of data supplied by the applicants. • To make such a comparison possible, the actually measured criteria need to be transformed to the measures of the above goals, taking into account the stochastic character of the initial expectations. • The evaluation process does allow to derive a subset of relevant needs assessment criteria from a larger set of criteria considered at the selection phase of the project.

22. Conclusionsregardingtheassessment of multicriteriaselectionmethod (2) A similar approach can be applied to assess the adequacy of the multicriteria decision-making methods that served to derive the final (ex-post) pilot sites’ ranking. The procedure can be outlined as follows: 1. Calculate the distance between the initial and final rankings (it reflects the efforts that would be needed to get the results anticipated, as seen ex-post.) 2. Define an aggregated value function, describing the quality of assessment and quality of performance of selected sites. 3. In case of repeating decision problems of the same kind, its outcomes can be used as inputs to the above-mentioned learning scheme. This approach can be applied to any problems involving an initial ranking, and the performance of N selected objects based on this ranking

23. Appendix: Internet transfer rates

24. Internet transfer rates

25. Internet transfer rates

26. Internet transfer rates

27. Internet transfer rates

28. Thank you for your attention! Furtherquestions and inquiries: ams@agh.edu.pl