Ikshana – AI/ML Trash Detection System Built With Python, Flutter, AWS

When it comes to hygiene and cleanliness in our homes, we prefer taking utmost care and incorporating maximum effort. But the one managing the cleanliness of a city requires a more organized approach to monitor each and every corner that needs to be cleaned. The Ikshana application we developed is based on a similar concept aiming to notice and address unattended garbage or dirty roads and apply a better disposal strategy.

Business Challenge


The client provides a public road cleaning facility by running garbage vehicles all over the city. The drivers usually ignore the roadside garbage and collect only the bags which are prepared by households containing their personally generated waste. The result was a wide range of areas that remained uncleaned due to the high number of unclaimed litters. The collected wastes lead to the emission of greenhouse gases, which increases the amount of carbon dioxide in the atmosphere, resulting in the heating up of our environment.

Solution


Not just the homes but our client concentrated on keeping the city and streets clean as well for which we designed an application in the name of Ikshana. The application administered three ends which includes the admin who manages the dashboard, the camera assignee and the cleaner.

Each and every garbage-picking vehicle associated with the client is assigned a camera with a unique ID that is saved with his/her name, vehicle number, and other necessary details on the database. The application for Camera Assignee developed leveraging the features of Flutter is server-based, which will also identify the objects over the road that are the source of the open detection systems for example a dust bean that is leaking and spewing waste, moving animals, filthy water upon that road, a dump, etc. Thus, if the application identifies any of the above-mentioned entities, it will automatically upload its pictures along with the place where it was taken including the latitude and longitude as well. Based on the identification of the vehicles that were used to make the catch and the person in charge of the vehicle a ticket will be generated automatically in the system to deal with it.

Now comes the admin who is the authorized person to access the tickets generated and can view it in the administrator dashboard. The admin has the power to appoint the appropriate person to clean the waste in the area, mark any trash as unhandled, and add the appropriate comments if necessary. The status of the ticket is completely accessible by the admin who can also close or reopen it if he still thinks the work may be improved.

The assignee is the one to whom the admin can assign the task of cleaning any area. He/ she has the opportunity to add remarks in the status update with photographs as soon as the ticket is changed to in progress. The application provides the option to change the status to approval after attaching an image and a note as proof.

160 Days

Duration

1280 Hours

Working Hours

Sep 2022

Live Date



Applied Technologies

Our Role in Client Success


The application emerged as an environment-friendly solution that streamlined the whole procedure of cleaning the streets by covering the areas which were once left untouched. The dashboard kept the admin, camera assignee, and the cleaning person on the same page and reduced the need for manual inspection. The auto-detection of images containing trash and the auto-filling of details of the camera assignee saved the time consumed in manually performing these tasks, thus promoting quick actions.


Thanks for watching!

Contact us today and we'll turn your ideas into success

Related Case Studies