Design and implement a mobile-based system to teach traffic signs to illiterate people

1 Introduction

Approximately 800 million individuals around the world are unable to read or write in their native language. Most of them are found in developing countries, especially in rural areas [1]. Some of these areas are suffering from infrastructural shortages, like electricity, water supplies, and so on. However, there have been huge improvements in technology in these areas, and most of the mobile providers are trying to attract customers there. The statistics have shown that the usage of smart mobile phones is about 43% with a significant percentage from developing countries [2]. This proves that smart phones play a big role in people’s lives. People who use smart phones use a variety of applications in addition to making phone calls and sending text messages. For example, smart phones are now frequently used to bring the information to their users. They have made a revolutionary change in helping people to obtain essential information through mobile-based applications. This has enhanced the way people communicate with society and the world [2]. However, most of the mobile-based applications use texts in their user interface [3]. Written text is a form of knowledge encoding which necessitates decoding for comprehension, and thus, it requires the decoder to be literate. As a result, most of the information and communication technologies, including mobile applications, target literate users while neglecting the world’s 800 million illiterate population.

Several researchers noticed this problem and tried to develop various systems to help illiterate people and improve their quality of life. For example [4], presented a web-based system to support learning for rural farmers, while [5] devised a visual phonebook, and finally [3], developed a website to help illiterate people in Pakistan in searching for a job. However, it is still challenging to find access solutions to the information and communication systems, especially for illiterate people [3]. Therefore, this research highlights this issue and presents an IT solution to help illiterate drivers in Iraq to access information and learn traffic signs through by using smart phones.

Having a driving license is essential for drivers, since it is the legal document which proves that a person is allowed to drive vehicles on the public roads. Before getting their driving license, the driver must pass a special test that is related to knowing traffic signs. Usually, passing this test is difficult especially for illiterate drivers, because they do not have access to traditional learning facilities like the traffic signs manual, as it requires reading the meaning of each sign. Moreover, the existing IT-based solutions like websites use only texts in the user interface which makes them accessible only for literates.

This paper aims to reduce the digital gap between literates and illiterates through building a mobile based application that helps illiterate drivers to learn traffic signs without needing to read or write. The app relies on voice instructions and is based mainly on graphical content (icons) that is designed especially for illiterate people. The app also provides a special test that simulates the real test taken by drivers before they can get their license. The test contains a number of random questions and multiple-choice answers, which enable the users to test their knowledge about the traffic signs.

2 Related work

Several studies have attempted to give solutions to illiterate people and reduce the digital gap between them and literates. Most of these studies have focused on the user interface design and interaction process to provide digital services for illiterates in different areas around the world. For example [3], conducted a study to provide access to information and digital services for illiterate people in Pakistan. The authors developed a website that helps illiterate people in the country to apply for an appropriate job without any human assistance. The design was based on the features of background audio and graphical content that are culturally relevant to Pakistani users. With the same purpose, a more recent study was conducted by [6] to digitize the job search for illiterate and semi-illiterate people in Bangladesh. Firstly, the authors conducted a requirement elicitation study to explore the considerations that should be taken before designing a mobile application that targets illiterates and semi-literates. The study came up with several recommendations, such as using the native language and relying more on voice, symbols, pictures and less on text in the user interfaces. Then, based on these recommendations, the authors developed an effective mobile application for searching employment in Bangladesh.

Similarly, in another study done by [2], the authors also developed a mobile app to support job seeking for illiterate people in Bangladesh. The app was mainly based on voices, symbols, intuitive icons, and touch-based interaction to explain the process of applying for a job. The Bengali language was used throughout the app interface and the voices.

Authors in [7] proposed a solution for the illiterate and semi-illiterate population in Pakistan using e-mail. The authors suggested a mobile application that contained several pages. Each of these pages reflected a step for building the profile of the illiterate user. The application interface was simple, rich in icons, and free of text.

In addition [8], presented a mobile health application to help illiterate and semi-illiterate pregnant women in Uganda. The app provides several functions, like an appointments reminder and making calls to health practitioners. The study suggested solutions for data input that could be utilized for illiterate people; for example, the login process is executed by entering depicted icons and a pictorial password. The app also relies on multimedia (video and audio) to display information about maternal health content and about reminders. Videos were used also by [9] as a solution for illiterate Moroccan people. The authors designed a useable E-government portal which served and was accessible by both illiterate and literate Moroccan populations. Table 1 illustrates the previous studies that have targeted illiterate or semi-illiterate people.

From all of the abovementioned studies, the author concluded that designing for an illiterate population must consider several factors, such as reducing or eliminating the usage of texts and utilizing vocal audios instead, all sounds must be in the local language, the focus should be on graphical content, conventional icons should be used, and finally, using videos could be helpful. However, it obvious that none of the previous studies had tried to give illiterates the necessary knowledge or to teach them through technologies. Therefore, this research attempts to fill the gap in the previous studies by developing a mobile application that targets illiterate people and tries to teach them using modern technologies like smart phones. Another critical issue is that there was only one study conducted in an Arabic environment [9]. However, this study gave the design considerations for building an accessible website interface for illiterates. Therefore, when developing mobile applications, it is important to understand the illiterate Arabic users and explore which design considerations are important to them.

3 Methodology

This study was conducted following the six steps of Design Science Research (DSR) suggested by [10] and used in [6, 11]. These steps are as follows: I- problem identification, II- objectives for the solution, III- design and development, IV- demonstration, V- evaluation, and VI- communication.

The first step aims to define the research specific problem and explain the solution. The illiterate people in Iraq do not have access to traditional facilities for learning traffic signs like the traffic signs manual, as it uses text for guiding and learning. Moreover, the existing ICT solutions, like some websites, rely solely on text in the user interface, which makes them inaccessible to illiterate people. As a result, the facilities for learning traffic signs in Iraq mainly target educated drivers and neglect illiterates. Therefore, an ICT (Information Communication Technology) solution for illiterate and probably semi-illiterate drivers is needed to help them learn the traffic signs without the need for reading texts.

The second step of DSR is to identify the objective of the solution. The goals of this research were to determine the design considerations for the illiterate population and to then build a mobile application that helps illiterate people in Iraq to learn traffic signs without assistance from others. There were some previous studies that suggested design considerations for illiterate people as it was mentioned in Section 2 of this study. However, these studies did not conduct in the Arabic culture and environment except one study by [9]. Although this study was carried out in Morocco (Arabic country), it was conducted on a web based application. Therefore, in this study we aim to identify the design consideration for illiterate people in Arabic culture and for mobile apps.

In order to define the design considerations for illiterate people and collect the user requirements, the author conducted a qualitative study with 25 participants. The results from interviewing these participants showed that using mobile applications for learning signs is preferable compared to using other tools. However, the participants were looking for an easy interface and interactive application. The requirement elicitation step resulted in several design considerations, which were used later in the application design and development.

The next step was to design and develop the application. The feedback received from interviewees in the previous stage was used to determine the application’s functionalities and the look of its screens. The following step was demonstration, where the developed app was used by several scholars, who gave their feedback on it. Then, the next step was the evaluation step where 25 illiterate people used the app and expressed their opinion about it. The data were collected and analyzed to measure the usability of the application. The final step was communication, where the study findings are reported. Figure 1 illustrates the steps of the research methodology used in this study.

Figure 1:

Steps of research methodology.

4 Application design and development

4.1 Application design

Based on the design considerations mentioned by participants in the requirement elicitation study, the author designed and built a mobile application for learning traffic signs based on serving illiterate people in Iraq. The goal was to design an application that was primarily based on graphical visual content and sounds with no text entry. Moreover, the app should be simple for users, so they will not need human assistance [15, 16]. Therefore, the app was developed with a simple visual interface and background audio that supports the users’ local language (Arabic). The main objective of the learning traffic signs application for illiterate drivers is to deliver the information about each sign using only sounds. The user can hear the sound by clicking on the “speaker” symbol. The same symbol is used on every page of the app. Then, the user can take a test that includes 20 random questions. The app uses sounds to read the questions and the answers options for users; the user can choose the right option by clicking on it and thus avoid any text entry. After answering each question, the app informs its users with their scores by speaking the score out loud. Figure 2 illustrates the design framework of the application.

Figure 2:

Learn traffic signs application framework.

4.2 Application development

Flutter, a cross-platform tool kit, was used to develop the app because it is an easy, high performance framework, and it supports both Android and IOS operating systems which makes the app available for a bigger number of users [17]. Flutter is based on Dart programming language, which has a large repository of software packages [18].

When the user launches the application, they will hear vocal instructions on how to use the app. Then, they can choose from “Learn”, to learn what each sign refers to, or “Test” to answer several questions about the signs. Users can identify which button is “Learn” or “Test” by clicking the “speaker” icon that is beside each one. In case of choosing the “Learn” button, a new screen that contains the sign groups will be displayed. In order to identify which sign group it is, the user should click on the “speaker” icon so they can select the one they would like to know more about. After the sign group has been selected, the app will launch another screen, which contains various signs that each driver must know in order to get their driving license. When the user has learned all the signs, they can take a test to make sure they have learned them and see if they are ready to take the real test or not. In this case, the user should click on the “Test” button and start answering the questions. Figure 3 demonstrates the activities flow of the application.

Figure 3:

Flow chart of the application.

4.3 Application implementation

Figure 4 depicts a few screens from the proposed application. As is shown, the entire application is developed with voices and graphical indications. The voice-guide is in the local language (Arabic) and is present on every page of the application interface. This will assist the users in understanding what to do in the current situation. The user is able to hear the explanation of each sign’s meaning by clicking the “speaker” icon, which is located on the right side of each sign. Obviously, the same symbol in the same location is used across the app in order to make it easier for users to memorize how to use the app. Additionally, the app provides its users with a test where the user can answer several questions by clicking on the correct option without the need to enter any text. Once the user has finished the test, the score is displayed on the screen with a background sound which tells the score. When the user passes the test (answers more than half the questions correctly), a green symbol appears to confirm that the user has passed. On the other hand, if the user has not passed the test and answers fewer than 10 questions, a red symbol is shown with “Try again” as a background voice message.

Figure 4:

Screen shoots from learn traffic signs application.

5 Application evaluation

The step after developing and testing the application was the application evaluation. A laboratory study was conducted with a number of illiterates to obtain feedback on the developed app.

5.3 Results

The data derived from the evaluation sessions were analyzed to measure the application usability in terms of three parameters: efficiency, effectiveness, and satisfaction [19, 20]. Table 3 summarize the results of evaluation sessions.

Table 3:

Summary of the evaluation results.

Evaluation criteria	Type of data	Task	Average	SD
Efficiency	Time to complete a task	Task1	30.2	±2.5
		Task2	52.04	±3.16
		Task3	5.48	±1.04
	Times the participant listened to the voice instructions	Task1	1.36	±0.63
		Task2	1.48	±0.7
		Task3	1.04	±0.2
	Times the participant asked for help	Task1	0.08	±0.276
		Task2	0.12	±0.331
		Task3	0	±0
Effectiveness	Number of taps required to perform a task	Task1	3.5	±0.714
		Task2	6.16	±0.8
		Task3	1.76	±0.59
User satisfaction	User satisfaction with the application	–	4.4	±0.7
	Easiness of the app	–	3.92	±0.8
	Using this application to learn traffic signs	–	4.52	±0.7
	Recommend this app to others	–	4.8	±0.4

5.3.1 Efficiency

According to [21], efficiency refers to when no time is wasted. In this study, three metrics were considered to measure the application efficiency: the time spent to complete a single task, the number of times the participant listened to the voice instructions, and how many times the participant asked for help. Figure 5 demonstrates the results of the three efficiency metrics.

1. The time to complete a task

Figure 5:

The results of three metrics used to measure the application efficiency.

The interface will be considered efficient if it requires less time to complete a single task [3]. In this study, the participants took an average of 30.2, 52.04, and 5.48 s to complete Task1, Task2, and Task3 respectively. The Standard Deviation (SD) for these three tasks were ±2.5 for Task1, ±3.16 for Task2, and ±1.04 for Task3. Figure 5(a) shows the average time to complete each task with SD values. The minimum time spent to complete Task1 was 25 and the maximum was 34. For Task2, the minimum time was 45 and the maximum was 57. Finally, Task3 took a minimum of 4 s and a maximum of 7. The number of participants who took longer than the average time to complete Task1, Task2, and Task3 was 6 (24%), 8 (32%), and 5 (20%) respectively. These results revealed that around 75% of participants had completed the tasks within the average time, which confirmed that most of the participants used the application efficiently.

1. Number of times the participant listened to the voice instructions

The application has Arabic voice instructions about how to use the app. If the participant listened one time to these instructions, this would be an indication of an optimal user interface design. On the other hand, listening several times to the instructions would indicate that the user interface was poorly designed [6]. To do Task1, seven of the participants (28%) in our study listened to the instructions more than once. For Task2, the number of participants who repeated the instructions was 9 (36%). However, only one participant (4%) went back to the voice instructions in order to know how to identify their score (Task3). This means about two thirds of participants were able to complete the tasks after listening to the voice instructions only one time. These results indicate that the voice instructions were clear enough, understandable to the participants, and supported the application’s efficiency by guiding the participants on how to perform the tasks.

1. The number of times the participant asked for help

The interface is considered efficient when the participants do not ask for help from the researcher to complete the tasks [6]. Our study findings revealed that only 2 participants (8%) needed to ask the researcher for help to complete Task1, and 3 participants (12%) needed help to complete Task2, while no one needed additional help on how to do Task3. These findings indicate that the participants found the application user interface to be clear and understandable. Moreover, the voice instructions helped the participants to learn how to use the application to perform the required tasks.

5.3.2 Effectiveness

Effectiveness is defined as the accuracy and success in achieving goals [21]. The parameter considered to measure the effectiveness of the learning traffic signs application is as follows:

1. The number of taps required to perform a task.

The minimum number of taps needed to do Task1, Task2, and Task3 was 3, 5, and 10 respectively. In the evaluation session, the participants tapped an average of 3.5, 6.16, and 1.76 times to perform Task1, Task2, and Task3 respectively. These results show that the participants took 0.5, 1.16, 0.76 more time than the optimal value of taps which was mentioned before (3, 5, 1 for Task1, Task2, Task3 respectively). Despites some of the participants needing slightly more taps for each task, the application is still considered an effective app. The biggest number of taps difference was for Task2 with 1.11, because this task is the most complex one in comparison with Task2 and Task1. Figure 6 explains the average number of taps required to perform a task.

Figure 6:

The results of measuring the application effectiveness.

5.3.3 User satisfaction

Satisfaction was described by [22] as “the expressed attitude towards a system by the user: emotions of contentment, pleasure, delight, relief and ambivalence”. In this research, the participants were asked four questions to measure their satisfaction with the application. The first question was ‘How satisfied were you with the application?’ The answers were classified according to a Likert scale ranging from 1 ‘Not satisfied at all’ to 5 ‘Very satisfied’. The second question was ‘Was the application easy to use?’ The answers ranged from 1 ‘Not easy at all’ to 5 ‘Very easy’. The third question was ‘Would you use this application to learn signs?’ The answers ranged from 1 ‘Not willing at all’ to 5 ‘Very willing’. The fourth question was ‘Would you recommend this app to others?’ The answers ranged from one for ‘Not recommend at all’ to 5 ‘Would strongly recommend’. The hypothesis was that the users would be considered to be satisfied if their average satisfaction ratings were equal to or more than 2.5 (the Likert scale mean value for satisfactory). The study results revealed that the users were extremely satisfied with the learning traffic signs application (mean 4.4). They also found the app was easy to use (mean 3.92). Moreover, the participants showed interest in the application and had a desire to use it for learning traffic signs (mean 4.5). Finally, they would strongly recommend the app to others (mean 4.8). Figure 7 illustrates the average results of users’ satisfaction on the application.

Figure 7:

The results of measuring users’ satisfaction on the application.

6 Discussion

In this paper, a mobile application was designed and implemented that would help the illiterate population in Iraq to learn traffic signs. The app was developed following the steps mentioned in the DSR method.

Firstly, the author conducted a requirement elicitation study to address the main design considerations that illiterates would like to add to or eliminate from the application design. The study revealed that one of the most frequent user preferences was that the app must focus totally on voices with as little text as possible. This issue was also addressed in the results of previous studies, such as [3, 9, 23]. These studies recommended mainly using audios in any application that target illiterates. However, these studies were based on web or computer applications. Therefore, this paper confirmed that using less text and depending on voices in the local language is also essential for mobile interfaces which are designed for illiterate people.

Secondly, after analyzing the data from the requirement elicitation study, the author found that focusing on graphical content is extremely important. The use of pictures, conventional or cultural symbols, and metaphors will assist illiterates to understand the user interface. The previous studies such as [3, 6], also recommended paying attention to these design principles. Thus, this study verifies these principles to develop mobile applications for illiterate people in Iraq. This research provides these principles which could be a template for designing interactive mobile applications for illiterate people even if these apps include learning and tests (questions and answers). In future, researchers could consider these principles for example, voice instruction and graphical content, when they decide to design a mobile app for the illiterate community in their countries.

On the other hand, the requirement elicitation study revealed that illiterate people expressed anxiety about using technology. As they had only little or no experience with using smart phones, they were concerned about using these devices; they thought that they would not be able to use the mobile app. The participants in the study by [6] expressed the same issue. However, after the proposed application had been developed and evaluated, most of the participants, who were all illiterates, showed an extremely competent performance doing the given tasks. They proved that based on the previously mentioned design principles (voice instructions, symbols, and local language), the developed learning traffic sign application was effective and efficient and achieved high user satisfaction. As a result, this study proves that developing mobile applications which target this group of people will increase their technological knowledge and reduce the gap between them and the people who have access to the information through modern technology.

Finally, the research has one limitation, which is the small sample size of participants who were recruited in the evaluation study. The number of participants recruited to give their feedback on the application was 25, which might be considered a small number. However, it was reported that 25–30 is an adequate number to reach data saturation in quantitative studies [24] and 12 participants in qualitative [25]. Therefore, the author thought that 25 participants were enough to address the research questions and obtain valuable results.

7 Conclusions

This paper presents a mobile application for illiterate people in Iraq to learn traffic signs. Two studies were conducted: the first, which was conducted to collect the user requirements in terms of the application design and their desire to use smart phones for learning, came up with several design recommendations for illiterate people, such as using voices throughout the application, focusing on graphic content with cultural icons, and using the local language.

The second study was conducted after the application was built in order to evaluate its usability and to obtain the users’ perspective on it. The evaluation results revealed that the developed application is efficient and effective, as the participants completed their given tasks within or less than the average time. The application had also gained user satisfaction, as they showed high interest in using it in the future. They would recommend it to others as well. These results proved that developing useable applications will motivate illiterates to use modern technologies more and will help them overcome their concerns about smart gadgets. Consequently, this will help to reduce the digital gap between literates and illiterates.