ATOM currently uses the integration with Captur.ai to implement real-time image analysis for parking compliance and verification. Captur is an AI-powered software system, trained on millions of end-of-ride images, providing immediate parking feedback.
Main benefits of using AI-powered image analysis:
Valid image detection - The AI module determines if the image is valid (the vehicle is present, good quality photo etc.), if the photo not valid user has to re-take the photo;
Get a live feed of AI-analyzed parking images to stay on top of bad parking cases (Possible answers - good parking, improvable parking, bad parking or insufficient info);
Use this data to redefine parking zones, penalize repeat “offenders”, and minimise bad parking behaviour;
Reduce the support ticket amount for your service.
Captur.ai helps to determine if:
scooter is visible on the end-ride photo;
the way the scooter was parked was acceptable or not.
How it works:
ATOM system takes the first photo of the end ride photo sequence from the app and sends it to Captur.ai - within 3-5 seconds (we show a spinner in the app while waiting) we receive an answer;
if all is good, we allow the user to end the ride;
if the photo is marked as not acceptable by Captur.ai (no vehicle in the image, for example), we show an alert message “Please retake a photo and make sure the vehicle is visible on the camera” and ask to retake the photo until Captur.ai says the image is alright. If the user tries 3 times to retake the photo and still can't submit the correct photo approved by Captur.ai, the system will allow the user to end the ride with the existing photo on the 4th attempt.
after the successful ride end, we send data to the dashboard and add a flag "good/bad/improvable parking".
How to enable it?
Reach out to the ATOM Mobility Support team to enable this feature for your service: help@atommobility.com
Limitations
Currently, Captur.ai supports scooters and bikes (e-bikes included).
Note there are extra costs related to the service.