Search⌘ K
Home/
Courses/
Grokking the Mobile System Design Interview

Dependency Injection and Modular Design in Mobile

Understand dependency injection and modular design, and how they interplay in mobile System Design.

We'll cover the following...

Mobile apps evolve fast, features are added, teams scale, and deadlines compress. But somewhere along the way, the architecture starts to fight back. We change one screen, and another breaks. Build times drag. Tests become fragile or non-existent. What was once a clean codebase becomes a maze of hidden dependencies, and tangled logic.

These issues directly impact product velocity and developer morale. Dependency injection and modular design address this by introducing structure, boundaries, and flexibility.

In this lesson, we’ll explore how dependency injection and modular design work together to create scalable, maintainable mobile architectures. You’ll learn how these patterns bring order to complexity, make features independently testable, and unlock the ability to scale your app.

What is dependency injection?

In mobile development, dependency injection (DI) is a key architectural principle that promotes loosely coupled components, testability, and improved code reuse. Unlike traditional object creation, where classes instantiate their own dependencies, DI delegates this responsibility to a centralized mechanism. This allows for a clearer separation of concerns.

At its core, DI is about providing an object with the resources it needs (its dependencies) from the outside, rather than having the object construct them itself. This leads to inversion of control (IoC): a central tenet in System Design where higher-level modules control the creation and wiring of lower-level components.

The three common forms of DI are mentioned below.

  • Constructor injection: Dependencies are passed in through the class constructor.

  • Field injection: Dependencies are injected directly into class fields.

  • Method injection: Dependencies are passed in through specific methods.

The following illustration provides an overview of dependency injection compared with traditional dependency initiation:

A comparison of manual dependency wiring vs. dependency injection
A comparison of manual dependency wiring vs. dependency injection

The left side of the illustration shows manual dependency wiring, where each component directly instantiates its dependencies using new. This approach leads to tight coupling, scattered instantiation logic, and poor testability.

The right side depicts dependency injection in action. Here, a centralized dependency injector manages the creation and provisioning of objects, adhering to the inversion of control principle. Each layer, application, activity, and ViewModel is injected with its dependencies based on well-defined scopesScoped dependency injection, or scoped-DI, is the practice of tying a dependency’s life cycle to a specific scope within an application. These scopes might include the application itself, an activity, a fragment, or a ViewModel. The injected object is created when the scope begins and destroyed when the scope ends, ensuring proper resource management and life cycle alignment.:

  • Application scope for app-wide singletons.

  • Activity scope for UI logic tied to the activity life cycle.

  • ViewModel scope for ...