--- title: Building a Formula E AI co driver on Android with ML Kit... description: Google Cloud engineers built an AI co-driver agent for a Formula E race car, running on a Pixel 10 Pro with Gemini Nano via ML Kit. A Raspberry Pi wired into... canonical: https://daily.dev/posts/building-a-formula-e-ai-co-driver-on-android-with-ml-kit-and-gemini-nano-8tv4ujtl2 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Building a Formula E AI co driver on Android with ML Kit and Gemini Nano | daily.dev og:description: Google Cloud engineers built an AI co-driver agent for a Formula E race car, running on a Pixel 10 Pro with Gemini Nano via ML Kit. A Raspberry Pi wired into... og:url: https://daily.dev/posts/building-a-formula-e-ai-co-driver-on-android-with-ml-kit-and-gemini-nano-8tv4ujtl2 og:image: https://api.daily.dev/og/posts/8tv4uJtl2.png og:image:alt: Building a Formula E AI co driver on Android with ML Kit and Gemini Nano og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Building a Formula E AI co driver on Android with ML Kit and Gemini Nano **[Android Developers](https://daily.dev/sources/androiddevelopers)** · 14 min read · 0 upvotes · 0 comments ## Summary Google Cloud engineers built an AI co-driver agent for a Formula E race car, running on a Pixel 10 Pro with Gemini Nano via ML Kit. A Raspberry Pi wired into the car's vehicle control unit reads CAN bus telemetry, downsamples it to 10 Hz, and streams it over MQTT to the phone. The driver says 'Hey Gemma' to ask questions like tire pressure or peak G-force, and the on-device model answers using live telemetry plus speech recognition, with Android's text-to-speech responding back. A cloud agent handles more complex queries via an agent-to-agent protocol. At Goodwood Festival of Speed, heat made the Raspberry Pi struggle and Bluetooth helmet pairing nearly failed, both fixed just before the run. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What model powers Google's Formula E AI co-driver demo on Android? The co-driver agent runs on Gemini Nano accessed through ML Kit, an on-device model running on a Pixel 10 Pro. It reasons over live car telemetry and driver speech input, using Android's Agent Developer Kit (ADK) as the application harness, with text-to-speech for responses and an agent-to-agent protocol for offloading complex queries to a cloud agent. _Developers building on-device agents can follow Gemini Nano and ADK coverage on daily.dev._ ### How did they get race car telemetry data into an Android phone in real time? A Raspberry Pi was wired directly into the Formula E car's vehicle control unit (VCU) over the CAN bus, decoding thousands of telemetry points per second like speed, acceleration, throttle and braking. It downsampled that data to about 10 Hz and published it over MQTT to a Pixel 10 Pro sitting in the cockpit, which the agent app consumed as one of its inputs. _Engineers wiring sensor data into mobile apps can track real-world telemetry pipelines like this on daily.dev._ ### What problems came up when deploying an AI agent inside a real race car versus a simulator? At Goodwood Festival of Speed, extreme heat caused the Raspberry Pi (not the Android phone) to struggle, requiring a UPS battery pack taped to its underside with sporadic recharging throughout the day. Separately, Bluetooth pairing between the driver's helmet and the app wasn't tested until race day and had to be hotfixed just before the car went up the hill. _Anyone shipping edge AI hardware can learn from these integration failure modes via daily.dev._ ## Similar posts on daily.dev - [How AI is boosting gender equality high-performance racing](https://daily.dev/posts/how-ai-is-boosting-gender-equality-high-performance-racing-uuvvij5f9) · The Register · 0 upvotes · 0 comments - [Bridging the Domain Gap: AI Race Coach built with Antigravity and Gemini](https://daily.dev/posts/bridging-the-domain-gap-ai-race-coach-built-with-antigravity-and-gemini-ffrdoq14l) · Google Developers · 1 upvotes · 0 comments - [Beyond the Chatbot: A Blueprint for Trustable AI](https://daily.dev/posts/beyond-the-chatbot-a-blueprint-for-trustable-ai-4g74onjot) · Google Developers · 0 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#android](https://daily.dev/tags/android), [#local-ai](https://daily.dev/tags/local-ai) [View this post on daily.dev](https://daily.dev/posts/building-a-formula-e-ai-co-driver-on-android-with-ml-kit-and-gemini-nano-8tv4ujtl2) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Building a Formula E AI co driver on Android with ML Kit and Gemini Nano","url":"https://daily.dev/posts/building-a-formula-e-ai-co-driver-on-android-with-ml-kit-and-gemini-nano-8tv4ujtl2","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/building-a-formula-e-ai-co-driver-on-android-with-ml-kit-and-gemini-nano-8tv4ujtl2"},"datePublished":"2026-10-30T20:31:13.846Z","dateModified":"2026-10-30T20:31:42.050Z","description":"Google Cloud engineers built an AI co-driver agent for a Formula E race car, running on a Pixel 10 Pro with Gemini Nano via ML Kit. A Raspberry Pi wired into...","image":"https://i.ytimg.com/vi/nV63eqjPeEI/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/nV63eqjPeEI/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Android Developers","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Android Developers","logo":"https://media.daily.dev/image/upload/s--VHoSw1TN--/f_auto/v1722583834/logos/androiddevelopers","url":"https://daily.dev/sources/androiddevelopers"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/building-a-formula-e-ai-co-driver-on-android-with-ml-kit-and-gemini-nano-8tv4ujtl2","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,android,local-ai","timeRequired":"PT14M","video":{"@type":"VideoObject","name":"Building a Formula E AI co driver on Android with ML Kit and Gemini Nano","description":"Google Cloud engineers built an AI co-driver agent for a Formula E race car, running on a Pixel 10 Pro with Gemini Nano via ML Kit. A Raspberry Pi wired into...","thumbnailUrl":"https://i.ytimg.com/vi/nV63eqjPeEI/sddefault.jpg","uploadDate":"2026-10-30T20:31:13.846Z","duration":"PT14M","url":"https://api.daily.dev/r/8tv4uJtl2","embedUrl":"https://www.youtube.com/embed/nV63eqjPeEI"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Android Developers","item":"https://daily.dev/sources/androiddevelopers"},{"@type":"ListItem","position":3,"name":"Building a Formula E AI co driver on Android with ML Kit and Gemini Nano"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/building-a-formula-e-ai-co-driver-on-android-with-ml-kit-and-gemini-nano-8tv4ujtl2#faq","mainEntity":[{"@type":"Question","name":"What model powers Google's Formula E AI co-driver demo on Android?","acceptedAnswer":{"@type":"Answer","text":"The co-driver agent runs on Gemini Nano accessed through ML Kit, an on-device model running on a Pixel 10 Pro. It reasons over live car telemetry and driver speech input, using Android's Agent Developer Kit (ADK) as the application harness, with text-to-speech for responses and an agent-to-agent protocol for offloading complex queries to a cloud agent. Developers building on-device agents can follow Gemini Nano and ADK coverage on daily.dev."}},{"@type":"Question","name":"How did they get race car telemetry data into an Android phone in real time?","acceptedAnswer":{"@type":"Answer","text":"A Raspberry Pi was wired directly into the Formula E car's vehicle control unit (VCU) over the CAN bus, decoding thousands of telemetry points per second like speed, acceleration, throttle and braking. It downsampled that data to about 10 Hz and published it over MQTT to a Pixel 10 Pro sitting in the cockpit, which the agent app consumed as one of its inputs. Engineers wiring sensor data into mobile apps can track real-world telemetry pipelines like this on daily.dev."}},{"@type":"Question","name":"What problems came up when deploying an AI agent inside a real race car versus a simulator?","acceptedAnswer":{"@type":"Answer","text":"At Goodwood Festival of Speed, extreme heat caused the Raspberry Pi (not the Android phone) to struggle, requiring a UPS battery pack taped to its underside with sporadic recharging throughout the day. Separately, Bluetooth pairing between the driver's helmet and the app wasn't tested until race day and had to be hotfixed just before the car went up the hill. Anyone shipping edge AI hardware can learn from these integration failure modes via daily.dev."}}]} ```