Being first to market with rideshare on CarPlay and Android Auto | by Michael Ramdatt

Our cross-functional development processBy: Aastha Bhargava, Jake Hercules, Erik Kamp, Michael Ramdatt, Nathan Van Fleet, Rex Lam, Kieran GuptaFor years, drivers have been clear about what they wanted: native Lyft support for CarPlay and Android Auto. They’ve made the request across social media platforms, through the app, and in feedback sessions with Lyft researchers. Until recently, Lyft couldn’t provide it due to our app architecture. So, drivers turned to DIY solutions to use infotainment. They used Google Maps, Apple Maps, and Waze, which required constantly switching between navigation and the Lyft app. Some drivers even resorted to using third-party software to cast their rideshare apps to their car screen.Lyft started out using third-party navigation to support rider and driver experiences. This approach allowed us to scale quickly in markets all across the United States. However, third-party integrations come with limitations—our unique rideshare platform needs, as well as the needs of riders and drivers, often went unmet. In 2019, we set out to build our own Mapping platform from the ground up: Lyft Maps — navigation built by Lyft for drivers on the Lyft platform. The launch of Lyft Maps in 2022 has enabled more seamless, delightful rideshare experiences — including the ability to finally launch support for CarPlay and Android Auto.We partnered with Apple and Google to design and build our in-car experience. Our integrations would not have been possible without the support of the platform developers. CarPlay and Android Auto operate using a “template framework.” New infotainment apps conform to the templates to ensure consistent, clutter-free experiences for both riders and drivers. Lyft’s Mapping and Driver Experiences teams worked closely with Apple and Google partners for more than a year to integrate Lyft Maps using their in-car frameworks. In some cases, our partners have helped create new templates to better support our unique rideshare platform needs.Driver and rider safety is top of mind in the design of CarPlay and Android Auto. The safety stakes are higher than normal for drivers giving paid rides to passengers. Drivers on the Lyft platform are strongly encouraged to use phone mounts to safely secure their phone while driving; unfortunately, some drivers choose to use their phones handheld. Car screens are the perfect solution — they’re usually much larger than the phone and are positioned safely. Drivers can take advantage of this in-car real estate to limit phone interactions, keep eyes up toward the road, and minimize unnecessary distractions.We were able to work within template constraints to design an experience that felt natural for rideshare. CarPlay and Android Auto automatically switch between different states: navigating to pickup, waiting for a passenger, navigating to drop-off, and heading home. Drivers no longer have to switch apps or use risky workarounds — rideshare just works.Important safety and rideshare-specific phone features have been ported over to the big screen, as well. Speed limits, live speed, and speeding warnings remind drivers to proceed safely. Bonus zones (regions that signal to drivers where demand is highest) appear natively on the in-car map. Rider avatars that indicate their live location show up automatically so drivers know where best to stop for a pickup.Lyft Maps has a relatively small team — building for the car required lean development (to start our team just had one engineer per platform). Earliest development began with explorations of the Android Auto and CarPlay framework, working side by side with Google and Apple to better understand feasibility. Engineers at our partner companies provided valuable guidance, helping us get to a working prototype within a couple of months. As complexity on the project increased, we transitioned the work from an exploratory phase to a major focus of our driver navigation pods.We iterated through the hiccups of creating an entirely new driver surface. Scaling to two navigation surfaces increases demand on the phone and has the potential to slow app performance. Rideshare navigation has to work all the time: safety, driver earnings, and rider time are all on the line. We hold a high bar for infotainment quality and have made ongoing stability a top priority.The process for testing these experiences on physical devices was a challenge. Apple and Google provide great emulation tools to test on a computer locally, but the process for testing on external devices and in-car displays proved to be an issue. Applications built on a developer’s computer cannot automatically run on physical devices due to various limitations. They have to be submitted to Google or Apple for provisioning before they can be used for live testing. Our team worked with our partners to create new processes and “internal test tracks” to approve special versions of our infotainment apps, unblocking early testing on physical head units.Our team used the product rigorously in the real-world to resolve performance and functional issues before sharing with drivers. This new experience was tested using simulation devices, on impromptu Friday roadtrips, and with real passengers in real Lyft rides. We were able to keep a tight feedback loop: finding and reporting bugs, fixing them in days (not weeks), and immediately cutting new versions of the apps for further testing. The experiences went from rough prototypes in May 2022 to driver-ready later that summer.Once ready, we gradually rolled out CarPlay and Android Auto to drivers via experiments so we could properly measure their impact. In particular, we looked at whether infotainment support aided in the switch from third-party navigation to Lyft Maps (it did!). We also paid attention to how riders responded — infotainment affected not only the driver experience but the rider experience as well. Riders could now see ride progress upfront and center.Various car manufacturers contributed to the complexity of infotainment through the introduction of additional software within their in-car displays or head-units, which added additional steps to enable infotainment. Through early driver reports, our team was able to field this early and send instructions to drivers to further help with the onboarding process.