At Autoware Foundation, we’re building a suite of open source AI models to revolutionize autonomous driving across the globe 🔥

By breaking down proprietary silos, ensuring full transparency, and accelerating time-to-market for OEMs and Tier 1 suppliers, open source AI models will redefine how the automotive industry innovates.

🔍 𝐌𝐞𝐞𝐭 𝐒𝐜𝐞𝐧𝐞𝐒𝐞𝐠
Today, we’re thrilled to unveil our first release: SceneSeg, the open source vision AI engine that processes raw camera feeds to spot every critical foreground object—whether it’s a car, pedestrian, animal, loose tyre… or something you never dreamed it would see.

✨ 𝐖𝐡𝐲 𝐒𝐜𝐞𝐧𝐞𝐒𝐞𝐠 𝐈𝐬 𝐚 𝐆𝐚𝐦𝐞-𝐂𝐡𝐚𝐧𝐠𝐞𝐫?
✅ 𝘖𝘱𝘦𝘯 𝘚𝘰𝘶𝘳𝘤𝘦 𝘍𝘳𝘦𝘦𝘥𝘰𝘮: Full transparency, community-driven improvements, and zero licensing barriers—anyone can build, customize, and trust the tech.
✅ 𝘎𝘦𝘯𝘦𝘳𝘢𝘭-𝘗𝘶𝘳𝘱𝘰𝘴𝘦 𝘝𝘪𝘴𝘪𝘰𝘯: Learns what “important” looks like without bespoke detectors for each object type.
✅ 𝘌𝘥𝘨𝘦-𝘊𝘢𝘴𝘦 𝘔𝘢𝘴𝘵𝘦𝘳𝘺: From pedestrians darting into traffic to stray tyres rolling across highways—SceneSeg sees the rare, unpredictable scenarios that matter most for safety.

🌍 𝐏𝐫𝐨𝐯𝐞𝐧 𝐖𝐨𝐫𝐥𝐝𝐰𝐢𝐝𝐞
Tested on roads from Singapore’s bustling streets to Los Angeles freeways, Istanbul’s cityscapes to Dubai’s desert highways—rain, snow, fog, low light… SceneSeg handles it all.

🔜 𝐖𝐡𝐚𝐭’𝐬 𝐍𝐞𝐱𝐭?
Stay tuned for our upcoming AI model introductions—more cutting-edge tools are on the horizon to power safe, scalable autonomy everywhere.

💬 𝐂𝐚𝐫𝐞 𝐭𝐨 𝐭𝐞𝐬𝐭 𝐨𝐮𝐫 𝐀𝐈 𝐦𝐨𝐝𝐞𝐥𝐬, 𝐨𝐫 𝐭𝐨 𝐣𝐨𝐢𝐧 𝐭𝐡𝐞 𝐝𝐞𝐯𝐞𝐥𝐨𝐩𝐦𝐞𝐧𝐭 𝐭𝐞𝐚𝐦?
We’re building these models in our Privately Owned Vehicle working group (free and open to anyone). Visit our LinkedIn announcement, comment on the post, and we will reach out to you for your onboarding.