BiconomyWallets

This project highlights the potential of taking real-time data from computer vision/AI to store product information and IoT data (i.e. location/timing) for storage into the blockchain. The value of the data is enhanced further by the potential of having issuers verify the data and for holders of the signed credential to reliably share the data due to its cryptographic proof. For example, imagine a supplier that pays only when the product gets to a certain port or location. By cryptographically signing the credentials, transportation responsibility and quality of goods, additional data can be used for decision-making and fraud/theft can be reduced. The same applications for supply-chain management can be applied to health or any industry where there is huge bureaucracy and poor centralized storage systems.

Also just as important, account abstraction using Biconomy is showcased. There is some real game-changers of how wallets can interact with the blockchain and what it means both for real-time applications of blockchain and the potential adoption. Some examples showcased in the demo include gas-less transactions, the ability to ensure that no native blockchain is required or is a barrier to transactions, sponsorship of a wallet balance that can be withdrawn over time, an access token to allow one sign-in for multiple transactions, and social login.

I used API for the computer vision. There is a dataset you can download and then you can utilize training methods to ensure the computer is trained to recognize strawberries.

For DID, I utilized Polygon-ID and Onyx JP Morgan's DID methodologies.

For everything account/wallet-related, I leveraged Biconomy tutorials on the web to produce the two examples (one where you can send variable amounts with no gas fees required, and the other where you have access to a pool of funds to send without re-signing).