Skip to main content Skip to navigation

Re-vue (Rear-View)

Re-vue is a smart cycling safety system that sees what you can’t, using radar and AI to detect danger from behind, alert riders and drivers, and help prevent accidents before they happen.

  • Final model mounted on bike seat to demonstrate real-world placement and user context.

  • Comparison of early AI and trained ML models showing improved accuracy using real cycling footage.

    Comparison of early AI and trained ML models showing improved accuracy using real cycling footage.

  • Re-vue in action!

  • Showcased final design models and fully working prototype at recent exhibitions and demonstrations.

  • Working early prototype using Raspberry Pi to run real time AI-based vehicle detection.

  • Exploded view showing enclosure design and internal components of the early prototype system.

What it does

Re-vue is a cyclist safety system using radar and AI camera detection to enhance rear awareness. It addresses blind spots, poor overtaking judgement and visibility, factors in 75 percent of serious collisions, to help prevent accidents before they happen.

Your inspiration

A cycling accident left my brother with a head injury after he was hit from behind. It completely changed how I view road safety. As a cyclist myself, I started looking into it more and learnt that thousands of similar rear impact incidents happen every year in the UK. Speaking with other cyclists confirmed just how exposed and anxious people feel when riding. Each year, 17,000 cyclists are seriously injured or killed in the UK, with over 75 percent involving poor rear awareness. As cycling grows, my brother’s experience gave me the personal motivation to design something that helps prevent this.

How it works

The device acts as another set of eyes for the cyclist, increasing confidence and helping prevent accidents before they happen. It combines radar and an AI powered camera to detect vehicles approaching from behind. The radar offers up to 150 metres of range, and the wide angle camera provides 160 degrees of coverage using a trained machine learning model. Together, they deliver accurate real time alerts with fewer false positives, making it ideal for both urban and rural roads. When a threat is detected, it responds instantly with LEDs, vibration and sound. The lights also brighten to alert drivers, improving cyclist visibility and offering layered safety for both rider and driver. The system runs on a Raspberry Pi prototype custom enclosure. Dozens of tests were run on ultrasonic, infrared, LiDAR and ToF sensors before selecting radar and camera fusion as the most reliable and adaptable combination. It becomes part of the rider, not just an accessory.

Design process

The process began with extensive research into cyclist safety, accident data, and rear collision risks. Interviews and surveys with cyclists helped identify the key pain points: lack of rear awareness, poor visibility, and slow reaction time. Early concept sketches explored smart helmets, modular lighting, and AI-driven systems. Foam models and ergonomic studies were used to test mounting options, visibility and form. I created multiple prototypes focusing on both form and function. Over a dozen sensors were trialled, including ultrasonic, infrared, LiDAR and ToF, but radar paired with an AI camera proved most effective for range and accuracy. Electronics were tested on breadboards before building the full working prototype using a Raspberry Pi 5, a fisheye night-vision camera, NeoPixel LED strip, and a vibration motor. I developed and trained an AI model using OpenCV and YOLO to detect vehicles in real time. User testing was carried out in day and night conditions to refine alert feedback and form factor. The casing was 3D printed and evaluated for weather resistance, heat dissipation, and usability. Each round of feedback shaped improvements in interface, mounting design and system reliability, leading to a compact, weatherproof prototype that feels integrated into the ride.

How it is different

What makes this design unique is its all in one approach to cyclist safety. It does not just detect vehicles , it combines long range radar and wide angle AI camera vision to accurately monitor what is behind in real time. First system of its kind to fuse radar and AI vision specifically for cycling safety. Most products use only one sensor or just alert the cyclist, but this system has better accuracy and reliability. By reducing false positives, it builds trust and confidence with the rider, avoiding the issue of alert fatigue. When a threat is detected, the rear LEDs also brighten to alert drivers, creating a multi-layer system. In the event of a crash, the camera acts as a dashcam and can trigger emergency contact protocols. All of this is housed in a compact, weather resistant unit that mounts easily to the bike. It does not feel like a bulky add on, but a seamless part of the ride. It offers prevention, awareness and response in one powerful system.

Future plans

Re-vue is now evolving into a startup, with further development, testing and validation underway. Recent showcases have sparked strong interest from industry professionals and potential investors. My next steps include refining the hardware, improving AI accuracy, and working towards full radar and camera fusion. I plan to launch a small batch for field trials, collect real user data, and explore options for crowdfunding or seed funding. My goal is to bring Re-vue to market as an accessible and life-saving product for cyclists everywhere.

Awards

Best Design for innovation, Keyshot Loves Award, Instuition of Enigneering Designers Award, Nomination for Design Intelligence Award.

End of main content. Return to top of main content.

Select your location