Deploy your apps to a supercloud in a few clicks

This Engineering Education program is supported by Section. Instantly deploy your GitHub apps, Docker containers or K8s namespaces to a supercloud.

Try It For Free

Basics of Android Workmanager

December 15, 2020

Background work has been a core part of Android application development for a long time. This is because it allows the execution of tasks without any interference with the user interface. However, implementing such functionality is not an easy task. One has to consider resources such as threads and when to start running the task.


Some tasks may even require running at time intervals. Several solutions have come up. A couple of these solutions are IntentServices and JobSchedulers. But they too come with their challenges like chaining tasks and setting constraints. This is where WorkManager comes in to help.

What is WorkManager

This is an AndroidX library that helps in running tasks asynchronously. It ensures that the work is done even if the user exits the app or the device restarts. It makes use of the existing job services to do the work. This means that it can support devices up to API level 14. WorkManager makes running background tasks easier and gives other awesome perks such as:

  • constraints: with WorkManager, you can easily set conditions that should be met for the task to run.
  • chaining: you can easily tie multiple tasks to run at the same time or one after another.
  • threading: like most AndroidX libraries, it comes with support for coroutines and rxjava for better thread management.
  • work execution: you have two options to define the execution of work. i.e. either once or periodically.

In a normal WorkManager setting, you have three parts:

  1. the work definition: this is where you define the task or rather, the job.
  2. the work creation: you create the job and set constraints to it. You may create it as a one time job or one that runs at intervals.
  3. queueing the work: you use the WorkManager instance to start/launch your job.


To comfortably follow through, you will need:

  • Android Studio installed.
  • Knowledge of Android application development.
  • Basic understanding of the AndroidX Room library.
  • Basic information of the Kotlin programming language.

Let’s get to it.

Step 1 - Setting up the project

To get the starting code for this tutorial, clone the project from GitHub and open it in Android Studio. Open the terminal in the IDE and run the following commands to rollback the project.

git checkout workManager
git checkout f1a4d683563ffe5c6eb00b3ea91353db0db2ca9a

After the gradle build finishes, add the following dependencies in the app-level build.gradle file.

kapt ''
implementation ''
implementation ''

// WorkManager
implementation ''

NOTE: Don’t forget to add kotlin-kapt as a plugin on the top of the gradle file.

Click sync to download the new dependencies and sync them to your project.

The application follows a basic MVVM architecture approach so go ahead and create a ViewModel class for your MainActivity class. You can go ahead an download this file or copy it’s code and add it to your application.

The code is in charge of generating random data for us.

The Room database is also set up in the db package. You can read more about Room in this article.

Step 2 - Defining our work

Create a new package and name it work. In here we will place our jobs or work.

Add a new Kotlin class named Work and add the following code.

class Work(context: Context, params: WorkerParameters): Worker(context, params) {


A worker class holds the tasks to be run by our job. The class receives Context and WorkerParameters as parameters. There is one method that needs to be implemented i.e. the doWork method.

    override fun doWork(): Result {}

This method runs our job on a different thread provided by WorkManager. The return type is ListenableWorker.Result which informs WorkManager the state of the job/work.

Go ahead and add the implementation of the work in the doWork method.

private val dao = AppDatabase.getDatabase(context).dao()
private val composite = CompositeDisposable()

override fun doWork(): Result {
    return try {
    catch (e: Exception){

override fun onStopped() {

First, we initialize the Room Dao and pass in the context available during the work execution. Then we create a CompositeDisposable that holds the disposables created during our job execution. We dispose it in the onStopped method when the work is done. To read more about RxJava in Android, you can go through this article.

We then use a try-catch block to execute our job, namely, to get a user and save them to our Room database. If any error occurs, the catch block returns a Result.failure() otherwise, we return a Result.success().

With that, our work is ready to go!

Step 3 - Scheduling the work

The next step is to register the work and set it rolling.

In our ViewModel class, we create a WorkManager instance and pass in a context.

// WorkManager instance
private val manager = WorkManager.getInstance(application)

Then we create our constraints. These are conditions that need to be met for our work to be done. Constraints range from network availability to device’s battery power. In our case, we will require the device’s battery level to not be low and for the device to be connected to the internet.

// Our work constraints
private val constraints = Constraints.Builder()

You can explore more constraints by checking the file available in the library.

NOTE: To access the library file and other core framework files, hold the ctrl/cmd key and click the Constraints object.

The only thing remaining is creating the work and setting the constraints. As mentioned earlier, there are two types of jobs in WorkManager.

1. One time work

This is a work/job that is run only once. It is created using the OneTimeWorkRequest class that takes one parameter, i.e. our worker class. We create one in our code and name it oneTimeWorker. We set constraints using the .setConstraints() method and pass in our constraints.

  // Define OneTime work
  private val oneTimeWorker = OneTimeWorkRequest.Builder(

2. Periodic work

Sometimes we may need the work to be run in intervals. In this case, we use the PeriodicWorkRequest class. This takes in three parameters, known as our worker class, the interval, and the TimeUnit.

Add the following code.

// Define Periodic work
private val periodicWork = PeriodicWorkRequest.Builder(, 15, TimeUnit.MINUTES)

We set our time interval to 15 minutes. This is the minimum interval defined in the documentation.

To start our work, we pass in the defined work into the WorkManager instance we created. We call the enqueue() method and pass in our work. Since we have multiple jobs defined, we can pass in a list of jobs as a parameter. Create a function called startWork() and add the following code.

fun startWork(){
    manager.enqueue(listOf(oneTimeWorker, periodicWork))

This starts the work for us provided the constraints are met.

Sometimes you may want to chain work, or start with one job and after completion, begin a second job. Then we can make use of beginWith() and then() methods.

Each of these methods receive one job as a parameter. They however receive work of OneTimeWorkRequest type only. So periodic work cannot be used in a chain.

Step 4 - Finishing up

Now that we have our work completely set up. You can go ahead and follow through the repository to check the UI setup.

Once the application starts, it calls the startWork function that starts our work. This starts both the oneTimeWork and periodicWork. The one time work completes immediately and adds one user to the Room database.

The periodicWork also starts by adding one user to the database and another user after 15 minutes. After 2 hours, the periodicWork will have added 8 users. On the first run the application should show 2 users. One from the oneTimeWork and another from the periodicWork.


With that, you have the basic information about WorkManager. As you can see, it makes it easier to schedule background tasks. You are able to run tasks like network calls even if the application closes.

A good use case could be backing up data on a different network like cloud services. This can be done when some constraints are met and you have the assurance of your work being done. Feel free to raise a PR or an issue with any updates.

Peer Review Contributions by: Peter Kayere