Software engineering process models for mobile app development: A systematic literature review

Highlights

• A comprehensive list of twenty process models for developing mobile apps.

• Usefulness to industry of the models based on rigor/relevance framework assessment.

• A synthesis of the models based on their Software Engineering areas and focus.

Abstract

Context: An effective development model can help improve competitive advantage and shorten release cycles, which is vital in the fast paced environment of mobile app development.

Objective: The aim with this paper is to provide an extensive review of existing mobile app development models.

Method: The review is done by following a systematic literature review process. Also presented is an assessment of the usefulness and relevance to industry of the models based on a rigor and relevance framework.

Results: 20 primary studies were identified, each with distinct models. Agile methods or state-based principles are commonly adopted across the models. Relatively little effort focuses on deployment, maintenance, project evaluation activities.

Conclusion: The review reveals that the contexts in which the identified models are intended to be used vary. This benefits practitioners as they are able to select a model that suits their contexts. However, the usefulness in industry of most of the models, based on the contexts in which the models were evaluated, is questionable. There is a need for evaluating mobile app models in contexts that resemble realistic contexts. The review also calls for further research addressing special constraints of mobile apps, e.g., testing apps on multiple-platforms, user involvement in release planning and continuous deployment.

Introduction

The popularity of mobile applications (mobile apps) has grown significantly with the advent of the trendsetting first generation iPhone by Apple. Since that release which was in 2007 (Islam and Want, 2014), smartphones have become virtually ubiquitous. Consequentially, companies have been trying to leverage this by trying to reach out to as many customers as possible through mobile apps. As a result mobile apps are a common feature of a variety of business domains. They range from simple entertainment leisure-consuming apps like games to apps within the safety-critical domain (e.g., mobile medical apps Food and Administration, 2013). Moreover, mobile apps contribute significantly to the lucrative mobile device market, which is estimated to be worth a trillion dollars (Fling, 2009, Perez). The market is however highly competitive and apps are often released at a high pace. There are software engineering processes that companies can leverage and improve app development processes and in turn gain a competitive advantage. The processes need to be tailored for mobile app development contexts (Wasserman, 2010). This review aims to systematically analyze existing peer-review literature and aggregate these tailor-made processes that model mobile app development, and also assess their relevance to industry.

Mobile devices are characterized as a portable device, viewed as a personal device by its users, and has a network connection (Firtman, 2010). Mobile apps are applications that run on these devices like smartphones (Maguire, 2013), and can be defined as, “a software application that can be executed (run) on a mobile platform (i.e., a handheld commercial off-the- shelf computing platform, with or without wireless connectivity), or a web-based software application that is tailored to a mobile platform but is executed on a server.” (Food and Administration, 2013). Typically, they can be downloaded from online application stores also known as App Stores, e.g., Apple’s App Store or Google’s Play for Android Apps.

In the context of this systematic literature review, we use the term “mobile app development model” to refer to a set of software engineering steps for developing mobile apps. The word “model” is used to refer to any process, approach, framework, method, model, set of guidelines, or life-cycle. Thus papers on, for example, coding or programming and development tools are excluded. The target papers, i.e., primary studies in our review, are those that propose, evaluate or validate a complete series of software engineering steps that are specific for mobile app development, but general and useful irrespective of the platform or devices. The assumption is that such models take into consideration aspects linked with mobile devices, which in turn constrain and distinguish mobile app development from general software development. Example constraints, which are also distinguishing aspects for mobile apps, are network connectivity concerns, hardware limitations (e.g., screen sizes and battery power), portability, reliance on sensors for many applications, user movement across multiple locations, and highly competitive markets with short time-to-market cycles (Wasserman, 2010, Fling, 2009).

Currently there is no indepth literature review of mobile app development models in peer-reviewed literature. The currently existing reviews only focus on models that are based on agile methods, e.g., Corral, Sillitti, Succi, 2013, Corral, Sillitti, Succi, 2013 and Flora and Chande (2013), or investigate other aspects linked to mobile apps, e.g., verification and validation of apps (Sahinoglu et al., 2015) and mobile software ecosystems (De Lima Fontao et al., 2015). Our review aims to identify mobile application models that have been proposed, evaluated and validated as reported in peer-reviewed literature, regardless of whether the models are based on agile methods or not. As part of the review, we also assess the usefulness to industry of studies on mobile app models using the rigor and relevance framework proposed by Ivarsson and Gorschek (2011). By only including peer-reviewed articles, the review in this paper provides state-of-art. We propose that future work should also investigate non-peer-reviewed articles, as well as processes in industry to better understand state-of-practice, and then compare with our findings.

Mobile app development has been in existence for more than two decades. However, our review only found 20 models and the first model was reported slightly over a decade ago, i.e., by Abrahamsson et al. (2004). Most of the studies in our primary study list had very low rigor and relevance scores. The implication is that it is difficult to argue for the usefulness of the models to industry. To summarize, the contribution of this study is as follows:

• The review provides an indepth overview of mobile app development models reported in literature, e.g., the intended use of the models and the software engineering principles or methods adopted in the models and adapted for mobile app development contexts;

• The review also shows mobile app development models that may be useful for industry and the contexts in which they are useful based on an assessment of the usefulness to industry of the evaluations performed on the models. This is done by assessing the transferability of the models to industry, and the trustworthiness of the benefits of the models, based on the type of evaluations performed on the models, e.g., if the evaluations are performed in realistic industrial settings.

The remainder of this paper is structured as follows. Background on mobile applications is discussed in Section 2, and this is followed by an outline of related work in Section 3. The steps followed in the review is described in Section 4. The results of each step are presented in Section 5. Validity threats are discussed in Section 6. A summary of the primary studies is presented in Section 7, then followed by the results and analysis of data extracted from the primary studies in Section 8. The results are discussed in Section 9. The paper concludes with an outline of the conclusions and future work in Section 10.