USING SIPRES – A FUSION OF THE REVISED SIMOS ’ PROCEDURE AND ZAPROS – IN THE ROAD ROUTE SELECTION

Road route decisions very frequently cause discussions and disagreement since they involve the number of stakeholders with competing interests. Before the construction of the road can start, the route for this road has to be determined, taking into account various facets, e.g. financial, technological, social and environmental ones. Such a situation can be described in the following way: the best possible choice must be made out of a finite set of alternatives (potential road routes) evaluated against a set of criteria. For this purpose different multicriteria decision aiding methods can be used, e.g. a novel tool called SIPRES. Its algorithm combines the key elements of the revised Simos’ procedure and the ZAPROS method. The method is transparent and easy to implement. On the one hand, it allows decision-makers to define their preferences simply and provides a straightforward but effective method for analysing the trade-offs between the alternatives using selected reference alternatives only (the ZAPROS-like approach). On the other hand, the revised Simos’ procedure applied in the method allows determining the cardinal scores for the alternatives. The purpose of this paper is to illustrate how the road route can be selected with the help of the SIPRES method, and to show thereby that this technique may be useful for solving such complex problems and may improve a decision-making process in certain situations.


Introduction
Road construction is a complex project which consists of many stages.Before the construction can commence the courses of the proposed route solutions have to be identified and assessed taking into account many different issues, e.g.functional, technical, economic, environmental and social ones.In many instances, several alternatives of the road route are examined (sometimes even over a dozen or tens).Information collected during this stage is used to determine the location and type of the road to be built [Górecka, 2013, p. 24].
Since the selection of a given road investment alternative is connected with certain financial, transport, ecological and safety effects, it is necessary to support the decision-making process with scientific approach.For typical economic and social impact assessment a widely used cost-benefit analysis (CBA) is sufficient.In other cases multicriteria decision analysis (MCDA) can be applied [Budzyński, Kaszubowski, 2014, p. 2405].
A few examples of such applications are briefly described in Table 1.
Other approaches that can be helpful in the analysis of the road investment alternatives and in choosing the most preferred one are, e.g.control lists and histograms as well as map, network and indicator methods [see: Podręcznik dobrych praktyk…, 2008;Szafranko, 2014].

TABLE 1. Examples of MCDA applications in the road route selection process
No.
Application Description (alternatives, criteria, approach) 1 Highway in the Kano area, Nigeria [Sunusi et al., 2015] This work presents a model developed by integrating Geographical Information System (GIS) with Analytic Hierarchy Process (AHP) and applies it to select an optimum highway alignment location which is economical and compatible with environment.
The aim of this study was to locate a suitable Least Cost Path (LCP) between two points that would pass major towns of Kura, Modobi and Kabo Local Government Areas (LGA) within the area under study.Three route themes were considered, namely engineering, environmental and a hybrid theme, and the last one turned out to be the shortest, the most economical and suitable.

2
East ring road of Warsaw, Poland [Określenie przebie-gu…, 2015] In this technical, economic and environmental study sixteen alternatives (eight basic ones and eight ones taking into account sub-alternative solutions) were considered.They were assessed using fourteen criteria, e.g.noise, destruction of habitats, number of wells in the area influenced by the project, nuisance of construction works, distance from Nature 2000 Training Ground Rembertów and level of acceptance by self-governments.Analytic Hierarchy Process (AHP) was applied to build a ranking of the alternatives considered.

3
Ring road of Malbork city, Poland [Budzyński, Kaszubowski, 2014] In the paper four alternatives of the Malbork orbital road were considered.Following criteria were taken into account in their evaluation: technology, transport, safety, environment, complementarity and the land availability.A two-step approach was used to select the most appropriate route: firstly, Analytic Network Process (ANP) was applied to determine the weights of the criteria, and secondly, the weights were transferred to the Analytic Hierarchy Process (AHP) to construct a ranking of the alternatives considered.Highway in Sinai Peninsula, Egypt [Effat, Hassan, 2013] In this paper Geographic Information System (GIS) tools were used to develop the least-cost path for a corridor to link three cities in a desert environment of Sinai Peninsula.Environmental and economic factors were integrated through a spatial multi-criteria model using Analytic Hierarchy Process (AHP).Three visions (routes) were taken into account: an engineering vision, an environmental vision and a hybrid one.A multi-criteria evaluation was used to compare these three routes and the hybrid route was finally recommended.E-763 highway entrance into Belgrade, Serbia [Marković et al., 2013] The subject of the analysis in this paper is the evaluation of two alternatives of the preliminary design of E-763 highway entrance into Belgrade by either the right or the left bank of the Sava River.In the assessment of these two potential solutions twenty criteria were taken into account and the ranking of the alternatives considered was obtained using VIKOR and PROMETHEE II methods.Road X, Somewhere [Górecka, 2013] Article presents the possibility of using multi-criteria decision aiding methods based on the outranking relation from ELECTRE and PROMETHEE families as well as methods belonging to the verbal decision analysis framework in the analysis regarding drafting of a road route.Input data, i.e. evaluations of five alternatives and weights of criteria, comes from [Biruk et al., 2007].In the assessment of the route solutions the following four criteria are taken into consideration: cost of realization, vehicle's average travel time, impact on the environment and safety of the travellers.Expressway 'Via Baltica', Poland [Jastrzębski, Kaliszewski, 2011] Study is focused on a real-life problem of choosing the most appropriate route for the expressway 'Via Baltica' from the Lithuanian border to Warsaw [see: Strategia rozwoju… 2008].Forty routes evaluated from the point of view of four criteria (traffic network criterion as well as economic, social and environmental ones) were considered in the article.Solver in Excel and the weighted Chebyshev function were used to find Pareto effective alternatives for the specified weights of the criteria.For selecting the most preferable route the use of an appropriate filter (e.g.cut-off values for the evaluations of the alternatives) within the mechanism of finding effective alternatives was suggested.Dublin port motorway, Ireland [Rogers, Bruen, 2000] This paper describes a practical application of the ELECTRE III method to ranking the various project options considered in a preliminary environmental evaluation conducted on the Port Access and Eastern Relief Route (PAERR) -a motorway proposed for Dublin City.

Expressway S6 (Lębork -Tricity ring road), Poland [Wybór wariantu…]
In this study the Analytic Hierarchy Process (AHP) was applied to select the most preferred road route from the environmental point of view.Eleven alternatives were considered, evaluated and compared using twelve criteria, e.g.noise, collision with protected species of plants, impact on protected species of animals, collision with the main ecological corridors, nuisance of construction works, impact on underground water, impact on soil and impact on material assets.10 Expressway S19, Poland [Raport o oddziały-waniu…] In this study the Analytic Hierarchy Process (AHP) was applied to select the most environmentally preferred road route for the expressway S19 from the border of Lublin and Subcarpathian Voivodeships to Sokołów Małopolski.Analysis was conducted separately for two parts of the expressway.In the case of the first part five alternatives were evaluated with respect to fifteen criteria.For the second part eight alternatives were assessed using nineteen criteria.
When it comes to MCDA methods, in the road route selection process AHP [Saaty, 2006;Saaty, Vargas, 1991] is frequently used (see: Table 1.).It is also recommended in document entitled Podręcznik dobrych praktyk wykonywania opracowań środowiskowych dla dróg krajowych [2008] in Poland, especially for choosing localization alternatives and environmental protection devices.According to this document, it can also be used in the selection process of technological and organizational alternatives as well as for environmental compensation [Podręcznik dobrych praktyk…, 2008, p. 164].Unfortunately, if the number of alternatives and/or criteria is high, then pair-wise comparisons on which AHP is based, may be really tedious and difficult for decision-makers (since large number of elements decreases the consistency of the comparisons conducted).In such a situation a recently developed technique called SIPRES [Górecka, 2015] can be applied.It has the following advantages: it allows decision-makers to define their preferences qualitatively, in a simple and effortless way, and it allows determining the cardinal scores for the alternatives.
The aim of this paper is to bring the SIPRES method closer to potential users and to show its usefulness in supporting decision-makers in the road route selection process.The article consists of an introduction, a conclusion and two sections.In the first section the SIPRES algorithm is described.The second section provides an illustrative example concerning the eco-challenging problem of a route selection for a road.

Overview of the SIPRES method
The acronym SIPRES stands for: Simos' procedure for Reference Situations.It is based on two methods: revised Simos' procedure [Figueira, Roy, 2002] and ZAPROS [Larichev, Moshkovich, 1995], and aims at obtaining a complete ranking of the alternatives with scores measured on a cardinal scale.The SIPRES method was introduced in 2015 [Górecka, 2015] as a continuation of the works on a tool for the verbal evaluation of the negotiation template connected with the MARS approach [see: Górecka et al., 2014;Górecka et al., 2016].Based on the original paper, in which two baseline methods were also presented, a detailed description of the SIPRES algorithm is given below. Let 1 is the set of all possible vectors in the decision space of n criteria; and 1 is a subset of X describing the alternatives considered.The SIPRES procedure consists of the following steps: 1.We determine the evaluation scale for each criterion considered in the decisionmaking problem.2. We prepare a set of blank cards and a set of cards with hypothetical alternatives (each with the best evaluation for all the criteria but one) as well as the ideal and anti-ideal reference vectors (with the best and the worst evaluations for all the criteria, respectively) and rank them from the worst to the best one.3. We introduce blank cards between two successive cards if necessary.The greater the difference between the evaluations of the alternatives, the greater the number of blank cards: a) no blank card means that the alternatives do not have the same evaluation and that the difference between the evaluations is equal to one unit u used for measuring the intervals between evaluations, b) one blank card means a difference of two units, two blank cards mean a difference of three units, etc. 4. We determine how many times the best alternative is better than the worst one in the ranking.5. We process the information obtained as in the revised Simos' procedure in order to obtain the normalized scores for the elements compared, i.e. to form the Joint Cardinal Scale (JCS).6.We substitute the evaluations in each vector describing the alternative considered in the decision-making problem by the corresponding scores from the JCS.For each alternative we define the distance from the ideal alternative using the formula: where ik p is the score from the JCS substituting the assessment of alternative i a according to criterion k f and max k p is the score for the best possible assessment for a given criterion.7. We construct the complete final ranking of the alternatives according to the distance values i L in ascending order.Processing the information in the way described in the revised Simos' procedure (mentioned in point 5 above) is as follows [Figueira, Roy, 2002, pp. 322-323]: 1. Let * n be the number of positions in the ranking, ' r e -the number of blank cards between the positions r and 1  r , and z -the ratio showing how many times the best element in the ranking is better than the worst one.We calculate: (2) (3) retaining six decimal places for u .Subsequently, we determine the nonnormalized score ) (r p for each position in the ranking: (5) where 0 0  e .We round these scores to two decimal places.If there are several elements in the same position r , all of them obtain the same score ) (r p ., with the required number of decimal places.The choice of the v elements, whose scores will be rounded, is performed using the following algorithm [Figueira, Roy, 2002, pp. 323-324]:

Let
1.For each element k g we determine the ratios: The key characteristics of the SIPRES approach are summarized below.

Application
Designed to elicit a sound preference relationship that can be applied to future cases; especially useful in the case of decision-making problems with mostly qualitative parameters and no objective model for their aggregation Decision-making problem More oriented to tasks with a fairly large number of alternatives, while the number of criteria is usually relatively smaller Decision-makers Does not require any special knowledge of decision analysis from the decision-makers; allows decision-makers to define their preferences in a simple and user-friendly way Methodology Combines the key elements of revised Simos' procedure and ZAPROS method to construct universal decision rules in the criteria space and then use them on any set of actual alternatives Source: own elaboration.

Illustrative example
The present study illustrates the application of the SIPRES method in transport planning decision-making.Its usefulness for decision aiding processes connected with route selection will be demonstrated by an example which concerns the problem of choosing the most environmentally preferred alternative of the road construction out of twenty five that have been identified at the stage of drawing up the project concept.
Let us assume that in this eco-challenging problem the following issues are discussed and taken into account in the alternatives' assessment: f 1 -negative project's impact on the inhabitants (noise, clearance of buildings, drinking water contamination), -f 2 -negative project's impact on the monuments and historical treasures (churches and chapels endangered, noise), -f 3 -negative project's impact on the landscape (endangered beauty spots, road slopes, areas visible from the road), -f 4 -negative project's impact on the environment (endangered trees, endangered habitats, intersections with the protected areas, endangered birds species from the Birds Directive, endangered plant species that are under strict protection).Evaluation scales for all the criteria considered have been defined linguistically.They are presented in Table 3. Table 4 provides the performance matrix for the twenty five potential routes considered and the four criteria used to evaluate them.
Table 5 presents the ranking of cards with hypothetical alternatives, determined by the decision-maker in accordance with steps 2 and 3 of the SIPRES algorithm.The ranking includes the alternatives with the best evaluations for all the criteria but one along with the ideal and anti-ideal alternatives.Additionally, the information required by step 4 of the algorithm is provided on how many times, in the decision-maker's opinion, the best alternative is better than the worst one.
Following step 5 of the algorithm, the information on decision-maker's preferences is processed to obtain the normalized scores for the elements compared, i.e. to form the Joint Cardinal Scale (JCS).The calculations conducted are shown in Tables 6-8.Source: own elaboration.
Tables 9 and 10 present the normalized scores for the hypothetical reference alternatives and the Joint Cardinal Scale respectively.The normalized scores reflect the scale of concessions required, when the ideal alternative is replaced by the alternative under consideration.Source: own elaboration.
Following step 6 of the SIPRES algorithm we substitute the evaluations in each vector describing the alternative by the corresponding scores from the JCS.For each alternative we define the distance from the ideal alternative and on this basis we build the ranking of the alternatives.The distances to the ideal alternative for each alternative considered as well as their ranks are given in Table 11.
Taking into account preferences determined by the decision-maker the most environmentally friendly road route is alternative a 10 .Straight after it, on the second and on the third place respectively, are alternatives a 4 and a 9 .In turn, the worst alternative from the ecological point of view is alternative a 21 .Source: own elaboration.

Conclusions
The SIPRES method presented in this article is an uncomplicated and functional technique that can improve the road route selection process, especially when the number of alternatives considered is large.In such a situation it is much less laborious and time-consuming than frequently used AHP.
Furthermore, this simple method requires the decision-makers to supply the basic preferential information only -they are able to operate with an intuitively interpreted card tool when defining preferences.Thanks to this technique we are able to determine the cardinal scale for the alternatives and build their ranking, in which no two alternatives will be incomparable.
Finally, it should be remembered that the applications of the SIPRES method are not limited to the complex transportation problems connected with the road route selection.It can be also applied in negotiation support to build a negotiation offers scoring system as well as in policy-making, strategic planning, R&D project selection and human resource management to order alternatives considered or to select the best one.
k g be an element in the position r , and ' s be the number of decimal places taken into account.We compute:

TABLE 9 . Normalized scores of the hypothetical alternatives
Source: own elaboration.