A Quick Recap of Google I/O

By Aftab Ahmad and Keane Quibilan

TribalScale’s Google I/O team is back and we’ve brought back plenty of new updates and exciting new features.

Here is a quick recap of what we learnt at Google I/O…

In line with the hype, the Google Duplex was demoed and announced as the latest update for the Google Assistant. The Duplex is a highly sophisticated version of the Google Assistant. It can make appointments, basic calls, reservations, etc., and sounds creepily human-like. You can watch a demo of the Google Duplex here.

One of the things Google promised at last year’s I/O was that they would double-down on machine learning and AI. This year, they announced the new Tensor Processor Unit (TPU). TPU 3.0 is so powerful that they added a cooling mechanism so that it would run properly, and it is supposed to be 8X faster than the previous version. With TPU 3.0, you can take a photo and then machine learning will identify what aspect is text and what isn’t. They had a demo with restaurant menus, for example: if you take a photo of a menu, the phone will detect the text, and you can search and translate the content live.

For Google Maps, AR and XR will be integrated so that an assistant, which was a cute fox in the demo, can direct you and help you find your way. A lens will also be integrated so that you can take a photo and know exactly where you are.

Google also introduced ML Kit in beta, which is a new software development kit (SDK) that will make machine learning much more accessible to developers using Firebase. Anyone with a basic math understanding will be able to use ML Kit. Since ML Kit is available through Firebase, developers will be able to access the platform and will be able to store image labels and measure processing latency, for example. ML Kit will allow developers to label images, scan barcodes, scan text, detect images, recognizes faces, and more! Getting started with machine learning has never been easier.


One of the more unexpected announcements was Google’s new initiative called “digital wellbeing.” The team said that around 70% of users talk of being addicted to their phone, they’re unable to put it down, and need an added level of control. In response, Google is introducing new features into Android P that will help with restraint. They are including a dashboard that will breakdown how much time an individual spends in each app. There will also be app timers that will let you set a specific amount of time to be spent on an app, once that time limit is reached the app will be greyed out. Another feature is “wind down mode.” Once it’s near bedtime, the phone will start dimming, colours will mute, the blue light will minimize, and eventually your phone will go to full greyscale before bed. Each of these initiatives are meant to help people move away from the digital world and regain time spent doing physical activity, socializing, etc. Soon, Google will be asking developers to integrate analytics into their app so that users know how much time is spent idling versus actual, engaged activity.

Project Nitrogen, a tool to help with test orchestration, running tests, and reporting, will be released in a few months. The neat thing with test orchestration is that you’re able to run tests uniformly, so whether it’s your local machine, CI server, or something else, they’ll all run the same way. Running and reporting tests is also cool with Project Nitrogen as it works with the new Android Test library, and depending on the platform (actual device, simulated, virtual, or just Java Virtual Machine) you use to run your tests, they’ll either be a unit or UI test.

Google announced their biggest update and revamp to the Material Design system since its initial release 4 years ago. They added more tools, including a Material Theme Editor, which is a plugin for sketch and allows you to build your screens in a unified way.

A handful of other announcements and releases were introduced at Google I/O. To list a couple more:

  • Gallery: a tool that Google has been using internally, allows users to upload sketch files, get feedback, and track changes.
  • Constraint Layout was updated and 2.0 will be published soon. The updated Constraint Layout allows for layout manipulation, post-layout manipulation, and has helpers which make it easier to build flexible and efficient layouts.
  • Support Libraries are improved with revised naming for packages and maven artifacts, new libraries, and the deprecation of the obsolete v4 and v7 libraries.
  • Jetpack: a set of libraries, tools, and architectural guidance for best practices helps developers make Android apps quickly and effectively.
  • Navigation component was added to Android architectures which simplifies, and allows a developer to deep-link while storing the state of the navigation.
  • Work Manager: an API that makes it very easy to specify and run specific asynchronous tasks — it’s opportunistic, backwards compatible, constraint-aware, queryable, and chainable.
  • Coroutines are production ready, however the final stable build hasn’t been made public yet.
  • Android Studio has added supports.
  • The list of updates and new features goes on, and on… However, many are still in beta. We’re looking forward to seeing how each of these updates work and what’s next from Google!


    Keane is an Agile Engineer and a Lead Android Developer at TribalScale.

    About the author

    Aftab is an Agile Engineer and Android Developer at TribalScale.

    Join our fast growing team and connect with us on Twitter, LinkedIn & Facebook! Learn more about us on our website.

    Visit Us on Medium

    You might also be interested in…