Google I/O 2016 Recap: Google Home, Android Instant Apps, Daydream

Contributors to this blog include Nathan Anaman, John Oberhauser, Noah Malmed, and Tyler Stewart

At this year's Google I/O Keynote, the Silcon-Valley giant displayed its vision for contextual technologies, the smart home, the next evolution of mobile apps, and much more. From Google Home to Android Instant Apps to Daydream, we pulled out the following key announcements for the enterprise.

Android Instant Apps

Google unveiled a new service called Instant App. Instant App allows users to experience an app without fully downloading it to their phone. Instant App allows users to temporarily load an app for a single use.

The key word here is ‘temporarily’. Once your done using the app, it doesn’t save on your phone.

The most compelling use-case they showed for this functionality is an app that allows you to pay for a parking meter. Now, a parking meter app isn’t the type of app that people tend to keep on their phone, so Instant App is really useful here because it allows users to have that mobile experience without having to commit to downloading a whole app. Expanding on this idea, we can imagine this type of service being really useful for something like a convention app that a user would heavily use for a weekend and never use again. Users tend to shy away from the commitment of downloading an app they know they won’t need in a couple of days. But with Instant App, you can build out an app and hopefully get a lot more people using because they don’t have to commit to the download.

One of the best parts of this service: it’s backwards compatible all the way back to JellyBean!

Google Home

The idea of a smart home has been around for some time, but has only recently become a reality. Google has made its largest step toward a smart home yet with the announcement of Google Home: a device that uses Google’s unmatched natural language processing to help you with everyday tasks. Users can set alarms, get recent news, and more. Google Home also gives the user control of multiple in home devices. This network, often referred to as the “Internet of Things” (IoT), will allow users to turn on lights, listen to music throughout the house, and play videos on their televisions using the groundwork already laid out with Google Chromecast and Chromecast Audio.

Google plans on working with developers to extend the possibilities of Google Home. Established businesses need to prepare to integrate their services. For example, users will expect quality restaurants to allow reservations to be placed seamlessly, or airlines to allow the purchasing of tickets with only voice commands. The possibilities of Google Home and IoT are endless and we can’t wait to see what people come up with. 

Allo and Duo

Google is releasing two new communication apps, Allo and Duo. Allo is a smart messaging app that adds many useful features, such as smart reply. Smart reply uses machine learning to give users quick replies that are based on previous conversations. Even more impressive is Allo’s ability to create a smart reply when users send photos to each other. Google’s publicly available computer vision API allows Allo to automatically identify the contents of an image and generate a conversational response to it. Allo also integrates the Google assistant (the same assistant the Google Home uses). Users will be able to do things like quickly search for restaurants and place reservations as a group, all within the app.

The second app, Duo, is a one to one video calling app. It will perform well on slow networks and feature end to end encryption. Probably the coolest feature of Duo is Knock Knock. It allows you to see the caller and prepare for the conversation even before you answer the call. Allo and Duo are set to be released this summer on both Android and iOS. 

Firebase

Ever since Google acquired Firebase back in 2014, it appears that they have been doing a major overhaul of the system. Before this announcement, Firebase was a backend service that offered real-time database, user authentication and site hosting capabilities. As of today’s announcement, in addition to all of its original features, Firebase now has analytics, messaging, storage, ad integration and a new pricing plan (hint: a lot of it is free!)


Firebase Analytics - A free and unlimited analytics engine. The analytics offered here are specifically designed for doing analytics on mobile apps.

Firebase Cloud Messaging(FCM) - Google’s new and free push notification platform.

Firebase Storage - Google’s new file storage service. Allows for client apps to upload and download large files such as images and videos.

AdMob integration - Firebase now integrates easily with AdMob in order to make ad integration in your app much easier.

The main take away of these improvements is that Firebase now offers a suite of services that are targeted to make the building of the backend of a mobile app much more streamlined. For most apps, Firebase can provide all of the backend functionality needed.

Machine Learning

Machine learning was mentioned frequently during this conference because it is the foundation of many of the new technologies Google is pushing forward with this year. Machine learning is the ability of software to become better at a task by processing data and recognizing patterns from it. Facial recognition, handwriting recognition, smart reply, and language understanding are all specialized applications of machine learning and Google has spent a lot of time working in these areas.

In keeping with tradition, Google is sharing this research with the developer community. Last year they released a paper on their FaceNet facial recognition algorithm, which boasts some very impressive accuracy numbers. Shortly after they released TensorFlow, a software library for building machine learning applications. TensorFlow is already used for things like Gmail and Google Photos, which enjoyed some spotlight during the Keynote.

This year Google is stepping up its game. They released a language understanding framework called SyntaxNet, and an English parser called Parsey McParseface. Leading up to Google I/O they’ve been working on a number of sleek cloud solutions and aggregated them under the Google Cloud Platform. This includes machine learning services such as computer vision and translations.

Google has left the door to artificial intelligence wide open by providing developers with the tools they need to develop their own solutions. At the same time the Google Cloud Platform provides an easy way to leverage machine learning for projects that don’t need customized intelligence solutions.

Daydream

Virtual Reality has been all over tech news for a while now (we talked about it back in December), but it's still clearly in the early stages. Oculus and HTC have both recently released their high-end Head Mounted Displays at high-end prices, and while some may get excited just thinking about donning the headsets and immersing myself in other worlds, the price tag isn't for everyone. That said, GearVR and Google Cardboard have been providing an affordable way for consumers to get their feet wet with VR, without breaking the bank. To step up their investment in widely-available VR, Google announced Daydream, an Android VR platform integrated into Android N. Using Daydream Home, users can launch VR compatible apps from inside VR, allowing users to move between applications easily. In contrast to the idea that VR headsets are only available to enthusiasts with money to spend on the hardware, the growing focus on Android devices as VR displays will get Virtual Reality into the hands of the average consumer. Google even plans to solve the issue of input in phone-based VR devices, with the announcement that a motion-tracked controller with a touchpad will be part of Daydream's required equipment. Google said they will be releasing their own version of this controller, but their main focus is on other hardware manufacturers to include this in their Daydream kits. This will go a long way toward addressing the fact that many Google Cardboard apps are, by necessity, passive experiences. Google Cardboard was nice (for an 8 dollar kit), but what Google is promising with Daydream is a commitment to VR as more than a novelty. 

Android N

For the upcoming Android release, Google has chosen to focus on performance, security, and productivity. These changes will further reduce the friction a user could experience while using Android and everyone that develops an app should take advantage of these new features.

For performance, Google created a brand new Just In Time (JIT) compiler that decreases app install time by 75% and reduces the storage needed for apps by 50% similar to Apple's App Thinning for iOS. Apps are getting larger because of the enormous amount of supported devices and both of these will help free up precious device space for users.

For security, Google has switched to using file based encryption and expanded its App Security Improvement Program by creating Safety Net. Safety Net checks eight billion apps per day to verify that they comply to recommended security standards. Google also will help app developers identify security risks within their app to ensure user data is not compromised. For example, Google has worked with various financial institutions to make sure they are using HTTPS properly.

Productivity also got a boost with expanded multitasking. Through data analysis, Google concluded that users only care about the last seven apps that they used and now only showing seven apps in the recent apps manager. Quick Switch allows you to double tap the home button to go back to the last app, similar to Alt+Tab on Windows or Cmd+Tab on OS X and multi-window allows users to enter split screen or utilize a picture-in-picture mode. Notifications also got a minor expansion with the ability to direct reply for messaging apps and long tap to adjust how future notifications of the same type are shown.

The full release will be available later this summer but a beta quality release candidate is available immediately. Also, Google is having trouble coming up with a name for the new release and is taking suggestions here.

Android Studio

Along with Android N comes a new version of their IDE, Android Studio 2.2. There are a bunch of new features that all of our Android developers are super excited to try!

Test Recorder - Arguably the most powerful tool added to Android Studio, the test recorder, allows you to generate automated UI tests just by using the app. The idea is that you hit record and run through a test case manually. As you run through the test case, each action you take is recorded as Espresso code. Once you are done with your test case, you’ll never have to manually run it again because you also just made an automated UI test. By enabling our QA team with this powerful tool, we can test faster and more efficiently. We find that the more vigorously we test our apps, the higher quality they tend to be.

New Layout Design - Android Studio is taking a page out of Xcode’s book by adding a new layout designer that automatically adds constraints to where you position widgets. This new layout designer is going to make UI design and creation much easier for developers.

To give the new version a try, head here for the download.

Wearables

Android wearables are getting an exciting change later this year. Right now, Android Wear devices can be connected to Wi-Fi to act on their own to some extent, while still relying largely on a tethered phone or tablet to power their core functionality. But with Android Wear 2.0, Google has announced that Android wear devices will soon support standalone apps, which do not rely on connection to a phone. They demonstrated a few, including a fitness tracker (no more putting your giant phablet in your pocket to track your jog) and a messaging app which allowed users to type by drawing out the characters one at a time, rather than being forced to use their voice in public, so no more looking like a crazy person talking to your wrist; now people will think you are a future-man operating your cybernetic implants using your watch. Not only will these apps support Wi-Fi and phone tethering, but also cellular connections (for wearables that include the hardware). They also showed off the ability to pin data from any app to any watch face, which is exciting for anyone whose app does more than tell battery life or weather. Android Wear 2.0 launches this fall.