Location Techniques for the Design of a Walking Aid Network for

Visual Impaired Students

Chefi Triki

Division of Engineering Management and Decision Sciences, College of Sciences and Engineering,

Hamad Bin Khalifa University, Doha, Qatar

Keywords: p-center Location Techniques, Walking Aid System, Visually Impaired Students.

Abstract: Universities may have students with visual impairments that require support and facilities in order to ensure

for help them adequate educational experience. Fortunately, many universities have already adopted particular

solutions for those students by providing supportive services and advanced technologies. However, other

campuses are still left behind and, thus, special solutions are still to be implemented. This position paper

represents a preliminary study in this direction for a case-study university campus in Oman. It attempts to

employ the location optimization techniques to design a walking aid network with rest lounges for students

who are blind or suffer from a visual impairment. The walking network will connect the several classrooms,

the administrative offices and the service structures that these students frequently visit during their academic

day. The walking network will be equipped by tactile tiles and has, thus, budget and execution time restriction

for its implementation. For this purpose, our approach will consist in selecting a subset of the available routes

to be used for the tactile tiles installation purpose and also in developing the project management technique

that allows to complete the project in the minimum span time. Preliminary results on the implementation of

such techniques in the Omani University campus will be discussed.

1 INTRODUCTION

Visually impaired students are persons who are

categorized as seeing very low and those who have

total absence of sight and need support to achieve

their daily tasks. The causes of visual impairment

differ from one case to another such as diseases

damage after birth that causes blindness or vision

impairment which cannot be cured by surgery or

medication. This paper deals with the orientation

and movement problem that the visually impaired

people are daily facing. It discusses how network

location techniques can help the visual impaireds

and how walking aid systems and appropriate

infrastructure can positively affect their life.

Designing special tracks can help the visually

impaired people through directing them from a

point to another inside the University Campus.

Moreover, this study is useful to the society; visual

impairment people will not use any further

resources to help them in their mobility and can,

https://orcid.org/0000-0002-8750-2470

thus, practice their life and be productive as normal

citizens.

The walking aid system is a path made from

tactile paving surfaces, used by the visually

impaired students to help them in conveying the

information, and in moving around the university

facilities. Nowadays, it is frequent to observe how

several visually impaired university students suffer

while moving around the universities’ buildings.

This happens in different campuses, and specially

in those established in developing countries, such

Oman. This fact often causes embarrassment for

this category of students who are forced to ask for

help or to reduce their movement.

Moreover, these students face often dangerous

situation as highlighted by Legood et al. (2020) who

reviewed several papers and reported that “people

with visual impairment are 1.7 times more likely to

have a drop and 1.9 times more likely to have multiple

falls in everyday living activities as compared against

the normally sighted persons. Moreover, the chances

of getting a hip fracture are between 1.3 times and 1.9

Triki, C.

Location Techniques for the Design of a Walking Aid Network for Visual Impaired Students.

DOI: 10.5220/0010344704590466

In Proceedings of the 10th Inter national Conference on Operations Research and Enterprise Systems (ICORES 2021), pages 459-466

ISBN: 978-989-758-485-5

459

times more prominent for those with visual

impairment.”

This work attempts to have a positive ethical

impact, by contributing in making the environment of

the university suitable for all types and varieties of

students. We will take Sultan Qaboos University

(Oman) as a case study. In its vast campus, the

colleges, facilities, roads, offices and car parking

areas are distributed far away from each other and

consequently the visually impaired students are

facing many difficulties to move around the campus

because of the total absence of any walking aid

system. Moreover, the streets/corridors of the

university are generally not suitable for blind and

visually impaired students who are forced to face

dangerous situations while moving from one building

to another or crossing the streets. The walking aid

system will, thus, reduce the difficulty of movement

of visually impaired students, reduce the risk they are

daily facing and speed up their movement around the

campus buildings. The network to be designed will

also include a number of rest lounges to be selected

among a set of potential locations for the exclusive

use of these students when moving from one building

to another or just to have some relaxing time.

The contributions of this paper consist in using

facility location techniques in order to design an

optimized path to allow visually impaired students to

freely move at any time without relying on other’s

help or exposing themselves to danger. We will use

for this purpose the p-center location technique in

order to design and optimize the walking paths and

optimally locate the rest lounges. We will also apply

the critical path method to schedule the activities

related to its implementation. Facility location

techniques have been intensively used in a wide range

of applications, such as waste management (Ghiani et

al., 2012), freight transportation (Triki, 2016),

railway hubs planning (Mokhtarzadeh et al., 2020)

and groundwater management (Triki et al., 2020). We

are confident that they can be a valid tool even in the

context of this application.

This paper is structured as follows. Section 2 will

be devoted to the literature review. Section 3 will give

a detailed description of the optimization models and

approaches used to solve the problem under exam.

Finally, Section 4 will conclude the paper.

2 LITERATURE REVIEW

According to the World Health Organization, visual

impaired people are defined as people who face a

limitation in their actions and functions because of the

visual system. Also, they have a low vision which

makes them unable to perform their daily routines

properly. The causes of visual impairment are

identified to be total darkness, glaucoma, and age-

related macular degeneration. The National Eye

Institute added more details by specifying that such

visual deficiencies usually cannot be treated even

through the use of glasses, contact lenses, medication

or surgery. There are three types of visual deficiency

depending on the vision level: (i) low visual acuity,

(ii) blind and (iii) visual impairment.

Tactile was invented in 1967 by Mr. Seiichi

Miyake. Yamauchi and Yamauchi (2003) claimed

that the Tactile Tiles (called also Tactile Ground

Indicators) were installed for visually impaired

people the first time in the Japanese city of Okayama

in March 1967. Their use started to expand quickly in

Japan specially after 1985 when the Japan National

Railways adopted them for the design of the stations

platforms. Furthermore, the Japan Road Association

(1985) gave insights on the different types of tactile

tiles that were manufactured and installed in Japan.

This latter document represented a guideline on the

installation principles of tactile tiles and shows also

the errors and mistakes that are usually made during

the implementation process. Besides the Japanese

experience, there are several countries that pioneered

the installation of tactile tiles, such as Korea,

Australia and New Zealand (see:

https://www.nzta.govt.nz/assets/resources/road-

traffic-standards/docs/rts-14.pdf). Another country

that defined rigid standards for the design and use of

tactile paving is United Kingdom. Indeed, in 1998 the

“Guidance on the use of Tactile Paving Surfaces”

(DETR, 1998) was published having the aim of

efficiently installing tactile surfaces to help visually

impaired pedestrians in finding their way and in

gaining important information which related to the

environment, danger, and location directions. The

guideline defines a blister paving designed to warn

the visually impaired pedestrians and give them a set

of important information. Tactile paving is divided

into two categories depending on the nature of its use:

(i) warning surface that include blister indicators for

pedestrian crossing points, corduroy hazard warning

surface, platform edge off-street warning surface and

on-street platform warning surface (Figure 1) and (ii)

amenity surfaces which include guidance path and

information surfaces (Figure 2).

There are many other recent attempts to propose

feasible solutions to guide visual impaired persons

that make use of modern technologies. For example,

Crandall et al. (2001) suggested the “Talking Signs”

system that uses infrared transmitters in order to

ICORES 2021 - 10th International Conference on Operations Research and Enterprise Systems

460

generate audible signage to label key environment

features. The spoken messages transmitted on a beam

of infrared light will be received by an ad-hoc device

hold by the blind or visual impaired person. The

authors have shown that this system greatly helps in

gaining in safety, and independence.

Figure 1: Example of corduroy hazard warning surface at

the top and bottom of stairs (DETR, 1998).

Another example is related to the use of sensors

embedded in a jacket that blind persons must wear in

order to be guided accordingly (Sourab et al., 2015).

Five sensors in total are installed with the aim of

detecting potholes or stairs, obstacles near head, and

barriers in front, right and left directions. Chai et al.

(2020a) have reviewed 30 articles that have dealt with

the identification of the recent studies on the smart

technologies developed for supporting the visually

impaired for their daily activities that involve both

obstacle detection and ground plane hazards detection

(see also Chai and Lau, 2020b).

Besides proposing the enabling technology for

assisting visually impaired persons, there is

nowadays a very intensive research work directed to

the navigation aspect in order to help these

pedestrians to plan their routes from an origin to a

destination location (see for example Kammoun et al.,

2010, Yusof et al., 2015 and the very recent review of

Nawaz et al., 2020).

From the above literature analysis, it is clear that

the research in this field has focused mainly on the

pre-installation and post-installation aspects of the

tactile tiles surfaces. To the best of our knowledge, no

work has dealt with the phase in-between that consists

in optimizing the design of the tactile tiles walking

aid system and the underlying network in a big

campus. Clearly, this problem arises when such a

campus has not been designed since its construction

to encompass tactile tiles and it will result, thus,

extremely expensive to cover all the buildings at this

stage. The focus is, thus, on identifying the building

(classrooms, offices, departments, etc.) that visually

impaired students frequently visit and connect them

with an optimized path having tactile tiles and

identifying suitable places for rest lounges.

Figure 2: Example of layout of guidance path surface at a

right angle turn (DETR, 1998).

3 NETWORK OPTIMIZATION

APPROACH

According to Hoffman and Ralphs (2013) many

practical optimization problems can be better

modelled and analysed by means of the powerful tool

of network optimization. This latter consists on a

graph representation in which a set of nodes are

connected with a set of links (that can be either

directed arcs or simply undirected edges). Network

optimization has been successfully used to solve

several logistics problems such as designing

highways, streets, integrated circuits and

telecommunication networks (see Ahuja et al., 1993).

In this study, we will make use of two specific

network models, namely the p-center facility location

and the critical path project scheduling models.

Additional techniques, such as shortest route, may be

needed at several stages of the design. The network to

be designed is schematically represented in Figure 3,

in which all the university buildings can be connected

directly or through any of the rest lounges to be

activated (among those potentially available).

Our general approach is based on the following

steps:

 Data Collection: data are mainly collected

through meeting visually impaired students in

order to understand their needs and to identify the

most important avenues they want the walking aid

system should reach.

Location Techniques for the Design of a Walking Aid Network for Visual Impaired Students

461

Figure 3: Network connecting the University buildings and

the rest lounges through feasible links.

 Network Analysis: identify the main buildings

inside the campus the visually impaired students

frequently visit and try to meet their needs by

considering them as nodes of our optimization

network.

 Network Algorithms: make use of the facility

location techniques in order to link all the

network’s nodes and to define, thus, the rest

lounges locations and the walking aid system

connecting all the buildings of interest at the

minimum paving cost.

 Project Management: finally, the University

administration needs to have insights on the time

span required to complete implementing the

walking aid network and the sequence of its

execution. For this purpose, we will apply the

critical path method to determine the project

scheduling. The critical path method will allow to

identify all the critical steps that will define the

priorities among the activities and the duration of

the project. The resulting outcome will be very

helpful for the University managers to assess the

best time to start executing the project and its

implementation phases.

After executing the first step, the resulting

outcome can be represented as shown in Figure 4,

characterizing the University campus, with the main

avenues frequently visited by the visually impaired

students (identified in red color) to be considered as

nodes of the network in the sequel. The resulting

network is clearly a multi-graph since there exist

several possible connections between many pairs of

nodes. However, we represented for clarity the

corresponding network (in Figure 5 in the Appendix)

as a simplified graph showing only a subset of

possible connections between nodes.

3.1 The p-center Location Model

The p-center model is applied to the decisions related

to the location of service centers (typical examples

are fire brigade stations and first aid centers) in the

event that it is necessary to ensure equity of treatment

to a set of nodes spread over a large area (the

university campus in our case). The goal is to locate

p logistic nodes (rest lounges here) in the area with

the aim of minimizing the maximum transfer time

between a node and the lounge closest to it.

The problem can be formally represented on an

undirected bipartite graph G (V, N, E) in which V is

the set where the lounges can be positioned, while E

is the set of edges identify all possible links between

the various nodes. Moreover, N indicates the set of

the destination buildings that are of interest to the

visually impaired students. Often, in these kind of

applications, sets V and N can coincide which means

that any destination building can be the location of a

potential lounge node as well. However, in the sequel

we will consider them different for the sake of

generality and to allow flexibility in adding or

removing any node to be (or not to be) considered as

a possible lounge location. To each edge (i,j)

∈ 

associated a cost d

representing the length of edge

(i,j), i.e. the distance from node i to node j. In the

particular case of a given building will be also

selected as a lounge location, then the building-

lounge distance will be equal to zero. The problem

will consists in minimizing the maximum distance

between any of the destination buildings and the

closest lounge to it. In this way, we are minimizing

the uncomfortableness of the visually impaired

students to reach a lounge from any of the buildings

of interest for them. The model’s decision variables

are as follows (see Daskin, 2013):

 



(

∀ ∈ ,∀ ∈ 

is 1 if destination building j

can be served by the lounge placed in node i.

Otherwise 



0;

 



(

∀ ∈ 

is 1 if a lounge is to be located at node

i, and zero otherwise.

Thus, the lounge location problem formulated as

a p-center model can be expressed mathematically as:

Min 

(1)

S.t.



∈

1 ∀∈

(2)









∈

 ∀∈

(3)





∈



(4)









∀ ∈ ,∀ ∈ 

(5)





∈



0,1



∀ ∈ ,∀ ∈ 

(6)





∀ ∈ 

(7)

ICORES 2021 - 10th International Conference on Operations Research and Enterprise Systems

462

The objective function (1) minimizes the

variables v which represents the maximum distance

between any destination building and its assigned

lounge. Such unrestricted variable v is defined

through constraints (3) as the highest distance of any

building-lounge assignment. Constraints (2) ensure

that every destination building is assigned to at least

one lounge. Constraint (4) will limit the number of

lounges to be opened to an upper bound value p. Such

parameter p is defined by the decision maker (the

University in this case). Set of Constraints (5) will

ensure that no destination building j will be assigned

to any node i if no lounge will be activated in that

node. Finally, Constraints (6) and (7) define the

binary nature of the decision variables.

The defined model (1)—(7) results to be a mixed

integer linear optimization program whose size

increases with size of the underlying graph. The

complexity of such the model depends to a big extend

on the value of the parameter p, as will be discussed

in the next subsection.

3.2 Solution Approaches for the

p-center Model

The specific case of 1-center location problem (i.e. in

which one single lounge must be located in the

network) the problem can be easily and efficiently

solved through the well-known algorithm of Hakimi

(Hakimi et al., 1978). However, for the general case

of p



2, the p-center model is known to be an NP-

hard problem (Ghiani et al., 2013). Therefore, the

solution of moderate sized instances can be

performed by solving the p-center optimization

model (1)—(7) by using exact solution methods such

as those implemented within the general purpose

commercial solvers Cplex,

Gorubi or Lingo.

Alternatively, specialized exact algorithms, such as

that developed by Daskin (2013), can be adopted and

applied. For example, the specific case of the network

shown in figure 5, having a limited number of nodes,

can be solved without difficulties with exact solvers

(a solution for p=2 is shown in Figure 6). However,

when the size of the problem (i.e. number of nodes)

increases, the instances become of large-scale and,

thus, the development of heuristic and/or meta-

heuristic approaches becomes necessary in order to

obtain a solution in a reasonable amount of time. This

will be certainly the case whenever the University

decides to cover not only part of the existing

buildings, but rather all the different structures of its

huge campus. There are several non-exact methods

that have been proposed in the scientific literature and

that can be employed to solve large-scale p-center

models (see the recent survey by Çalık et al., 2019).

Examples of such heuristics include the one developed

by Mihelic and Robic (2005) who used a

decomposition approach that consists in solving a

finite series of minimum dominating set problems.

Also, Bozkaya and Tansel (1998) developed a heuristic

method in which the p-center model is expressed in

terms of the spanning tree problem. Several meta-

heuristic approaches have been also developed to solve

the p-center problem, such as multi-start local search,

tabu-search, variable neighborhood search and

memetic genetic algorithm (see (Mladenovic et al.,

2003 and Pullan, 2008). The other direction of

investigation towards solving p-center instances

consisted in identifying lower and upper bounds that

were combined with Daskin’s algorithm in order to

design efficient a multi-level meta-heuristic approach

(Salhi and Al-Khedhairi, 2010).

4 CONCLUSIONS

The aim of this position paper is to apply network

optimization techniques that will help in designing a

walking aid network for visual impaired students in a

big University campus. The network will also involve

the location of a set of rest lounges to be used by the

visually impaired students during their movements

from one building to another. We suggest to use here

the p-center location method to solve this problem.

However, after the coding and carrying out an

experimental study it should be easy to understand if

the results of these techniques can be considered as

satisfactory, or more specialized approaches should

be adopted. Moreover, if the number of

offices/building increases, then solving exactly the p-

center model is not tractable and adopting heuristic or

metaheuristic approaches becomes necessary. This is

a preliminary work that should be completed.

ACKNOWLEDGEMENTS

The author have carried out part of the work related

to this article while he was affiliated with Sultan

Qaboos University. The support given the students,

staff and the administration of Sultan Qaboos

University is highly appreciated.

Location Techniques for the Design of a Walking Aid Network for Visual Impaired Students

463

REFERENCES

Ahuja, R. K., Magnanti, T. L., Orlin, J. B., 1993. Network

Flows: Theory, Algorithms, and Applications, Prentice

hall.

Bozkaya, B. and Tansel, B., 1998. A spanning tree

approach to the absolute p-center problem. Locat Sci

6:83–107.

Çalık, H., Labbé, M., and Yaman, H., 2019. p-Center

problems. In Location science, pp. 51-65. In Laporte,

G., Nickel, S., & Saldanha-da-Gama, F. (Eds). Location

Science. Springer, Cham.

Chai, A.B.C., Lau, B.T., Pan, Z., Chai, A.W., Deverell, L.,

Al Mahmud, A., McCarthy, C. and Meyer, D., 2020a.

Comprehensive Literature Reviews on Ground Plane

Checking for the Visually Impaired. In Technological

Trends in Improved Mobility of the Visually

Impaired (pp. 85-103). Springer, Cham.

Chai, A.B.C. and Lau, B.T., 2020b. Exploring the outdoor

performance of a LiDAR-based ground plane checking

system for the visually impaired. EAI Endorsed

Transactions on Pervasive Health and Technology,

6(23).

Crandall, W., Bentzen, B. L., Myers, L., Brabyn, J., 2001.

New orientation and accessibility option for persons

with visual impairment: transportation applications for

remote infrared audible signage. Clinical and

experimental Optometry, 84(3): 120-131.

Daskin M.S., 2013. Network and discrete location: models,

algorithms, and applications, 2

Edn. Wiley, Hoboken.

DETR, 1998. Guidance on the use of Tactile Paving

Surfaces. Department of Environment, Transport and

the Regions. United Kingdom. Available at:

https://assets.publishing.service.gov.uk/government/up

loads/system/uploads/attachment_data/file/289245/tact

ile-paving-surfaces.pdf (last accessed on 12 Nov.

2019).

Fu, L., Sun, D., Rilett, L. R., 2006. Heuristic shortest path

algorithms for transportation applications: state of the

art. Comp. & Operations Research, 33(11): 3324-3328.

Ghiani, G., Laporte, G. and Musmanno, R.,

2013. Introduction to logistics systems management.

John Wiley & Sons.

Ghiani, G., Laganà, D., Manni, E. and Triki, C., 2012.

Capacitated Location of Collection Sites in an Urban

Waste Management System, Waste Management,

32(7), pp. 1291–1296.

Hakimi, S.L., Schmeichel, E.F. and Pierce, J.G., 1978. On

p-centers in networks. Transportation Science, 12(1),

pp.1-15.

Hoffman, K. L., Ralphs, T. K., 2013. Integer and

combinatorial optimization. In Encyclopaedia of

Operations Research and Management Science, 771-

779.

Japan Road Association, 1985. The Guideline for Tactile

Tile Installation for Visually Impaired Person (in

Japanese).

Kammoun, S., Dramas, F., Oriolaand, B., Jouffrais, C.,

2010. Route Selection Algorithm for Blind Pedestrian.

International Conference on Control, Automation and

Systems KINTEX, Gyeonggi-do, Korea.

Legood, R, Scuffham, P & Cryer, C., 2020. Are we blind to

injuries in the visually impaired? A review of the

literature. Injury Prevention, 8(2): p. 155.

Mihelic, J. and Robic, B., 2005. Solving the k-center

problem efficiently with a dominating set algorithm. J

Comput Inform Technol 13:225–233.

Mladenovic, N., Labbé, M. and Hansen, P., 2003. Solving

the p-center problem with tabu search and variable

neighborhood search. Networks 42:48–64.

Mokhtarzadeh, M., Tavakkoli-Moghaddam, R., Triki, C.,

and Rahimi, Y., 2020. A hybrid of clustering and meta-

heuristic algorithms to solve a p-mobile hub location-

allocation problem with the depreciation cost of hub

facilities. To appear on Engineering Applications of

Artificial Intelligence.

Nawaz, W., Khan, K.U. and Bashir, K., 2020. A Review on

Path Selection and Navigation Approaches Towards an

Assisted Mobility of Visually Impaired People. KSII

Transactions on Internet and Information Systems

(TIIS), 14(8), pp.3270-3294.

Pullan, W., 2008. A memetic genetic algorithm for the

vertex p-center problem. Evol Comput 16:417–436.

Rygorash, O., Zhou, Y., Jorgensen, Z., 2006. Minimum

spanning tree based clustering algorithms. In 18th IEEE

International Conference on Tools with Artificial

Intelligence, ICTAI'06., 73-81.

Salhi, S. and Al-Khedhairi, A., 2010. Integrating heuristic

information into exact methods: the case of the vertex

p-centre problem. J Oper Res Soc 61:1619–1631.

Sekiguchi, H., Nakayama, H., 2002. On a history and a

present circumstances of walking aid for persons with

visual impairment in Japan. 5th International

Conference on Civil Engineering, Manila, Philippines.

Sharma, P., Khurana, N., 2013. Study of optimal path

finding techniques. International Journal of

Advancements in Technology, 4(2): 124-128.

Sourab, B.S., Ranganath Chakravarthy, H. S., Dsouza, S.,

2015. Design and implementation of mobility aid for

blind people. International Conference on Power and

Advanced Control Engineering, ICPACE-2015.

Triki, C., 2016. Location-Based Techniques for the

Synergy Approximation in Combinatorial

Transportation Auctions, Optimization Letters 10 (5),

pp. 1125-1139.

Triki, C., S. Zekri, A. Al-Maktoumi, and Bazargan-Lari, M.

R., 2020. Optimal Location of Wells for Storage and

Recovery of Surplus Desalinated Water in Coastal

Aquifers. Groundwater, 58(5), pp. 831-841.

Yamauchi, S., Yamauchi, Y., 2003. Tactile Ground Surface

Indicators for Blind Persons. National Rehabilitation

Center for Blind Persons

http://www.rehab.go.jp/english/whoclbc/pdf/E13.pdf.

Yusof, T.S.T., Toha, S.F., Yusof H. Md., 2015. Path

Planning for Visually Impaired People in an Unfamiliar

Environment Using Particle Swarm Optimization.

Procedia Computer Science, 76: 80 – 86.

ICORES 2021 - 10th International Conference on Operations Research and Enterprise Systems

464

APPENDIX

Figure 4: Campus map with the buildings of interest for visually impaired students highlighted in red color.

Figure 5: Resulting network (a simplified graph showing only a subset of possible connections between nodes).

Location Techniques for the Design of a Walking Aid Network for Visual Impaired Students

465

Figure 6: A solution representation with 2 rest lounges (placed in nodes 3 and 6).

ICORES 2021 - 10th International Conference on Operations Research and Enterprise Systems

466