the mobile App to improve scalibility
The educational purpose of the Parki ar minda project is to raise awareness on waste reduction and its more responsible treatment in Georgia. We are committed to improve both the quality of our eco-education and also the quantity of it to be provided.
THIS IS A SHORT SUMMARY. FOR THE COMPLETE DOCUMENT PLEASE REFER TO THE PDF FILE
Here, in the digital department, we take such the quantitative understanding of knowledge quite literally. We count it as the amount of population thathas actually adopted our sustainability guidelines in the daily life. In other terms the faster we can scale up our eco-taxi and other services, the more will be the educational outcome. The more people will have the reducing and recycling procedures “up and running”on their minds – the bigger is the knowledge produced for the eco education. This knowledge, the bigger it becomes the better it spread from our existing audience to the new one. To partially automate and improve knowledge distribution among the publicand our potential customers we develop the waste recognition machine learning model, and it needs your help!
The faster we can scale up the eco-taxi and other services, the more will be the educational outcome
Wouldn’t it be wonderful to automate the compound recognition for wastes, both in industry and households? It saves the need to explain how exactly one needs to recognize locally branded plastic packages one from another. Instead of manually following the eco-taxi sorting guidelines by every user, let the app do the recognition for us?
Unfortunately the artificial intelligence machinery involves a big flow of data which isn’t secure.
In order to adopt technology safely humans also need to control the impact of their data, which can serve targeted commercials and other influences on public, including scam and cyber-attacks. Common practice of including opaque third party infrastructure for seamlessly extracting the data within the app, puts users in vulnerability before intruders and the AI itself, which when at scale, may act dangerously misleading by its own.
Along with the ecological footprint we need to reduce the insecure data footprint, while still increasing the automaton potential, since we believe that progress persists to be the human’s only resilient hope for survival.
We guarantee your data will not fall into random hands just like we guarantee the recycling is real. Customers’ pictures of the recyclable wastes and related data will be never sent to the internet, It will be transported physically from users to our facility by the same route the waste goes and kept always encrypted and “air-gapped” i.e. constantly offline..
With this we are eager to set an example of the height standard for the data security in edge-computing available to the small business and communities. Environmental thinking has inspired us to seek a particular elegance in taking care of personal data, potentially containing insights about consumption, habits, and important events in personal people’s lives.
Please help us to obtain the necessary data explicitly by picturing and labeling objects in your recyclables sets. We will then collect and transport your pictures safely, keep them encrypted and offline until they are used for the waste recognition App (and nothing else) under our careful supervision.
No internet, no corporate servers needed to operate our smart solution, we are building it entirely on the basis of the previous research and the responsible support from our community.
If you want to help us with an autonomous waste recognition project, please download the What I Recycle for Android, and iPhone.
With the success of the compound recognition safe AI program, Eco-taxi will be able to scale up quickly by saving the cognitive friction when instructing new customers, and also show the significant sustainable opportunities for other recyclers and responsible businesses around the world.
Arkady Krakovsky
Vake Garage: 6 Mtskheta lane
Saburtalo Garage: Al. Kazbegi, 34a
Chugureti Garage: 19 Queen Tamar Ave
