Every year, hundred of thousands tourist lose their valuables and get scammed by the fake authorities. They could be in the form of tourist guides, government authorities, or private services. Having a distributed trust network can solve this trust problem by gathering authorities and tourists on the blockchain based self-sovereign identity platform. Authentication is further improved by the usage of state of art deep learning based solution implemented on cloud for scalability and easy rollout for future.

What it does

It brings the authorities(government, private, or local ), issuers(identity verifies), and the verifies (tourists) on the immutable secure blockchain based network. This network is implemented on the open source based self-sovereign identity platform, which is one of the most stable, trusted, and popular framework. Identity of the authorities is published on the block chain network and is being verified by the issuers based on request from verifier. To improve the usability and easy rollout, we have used 98% accurate machine leaning algorithms to make it happen without hassle.

How we built it

It is built on the python based open-source framework for self-sovereign identity platform named Hyperledger Indy, python based neural network library named Tensorflow, python based computer vision library named OpenCV and high compute is obtained from Microsoft Azure Cloud platform. The whole solution is working from end to end as demonstrated during the presentation.

Challenges we ran into

During the time-limited stress period of Hackathon, although we made fun at every moment, we came across following challenges,

  1. Network and VPN issues.
  2. Setting up GPU and Neural Network libraries on Cloud.
  3. Running out of resources while running the neural network, computer vision, and synchronisation for blockchain based solution.
  4. We modified our idea a bit while developing and the focus was to develop something to get the most flexible consumer experience.
  5. Stumbled while playing with blockchain based API and for making a logical workflow for different API calls.

Accomplishments that we're proud of

We had lot of fun, met cool people, got bundle of help from Hackathon team, and the most important, we are super proud of our end to end running prototype running on Cloud and it could be used to technically validate our idea. Apart from that, we are also proud of 98% accurate deep learning based model, efficient workflow of the Indy framework API calls, and computer vision based preprocessing.

What we learned

Since we all have almost disjoint backgrounds, we learnt a lot from each other. Martin is from corporate world with more than five decades of experience, Farhan is software developer with craze for blockchain based solutions and Muhammad was experience with machine leaning and cloud based deployments. Presentations delivered by the hackathton partners were super helpful.

What's next for Trust Travel

On our roadmap, we have planned to promote this idea more on networking sessions and authorities. We will try to validate the integration of the technical prototype. Since, we all are from different locations, so we have planned to met and develop more features based on feedback from the iterations. In 6 months timespan, Muhammad and Farhan will be graduating from their Masters program, and they have planned to get onboard with Martin as their business partner to continue working on the idea full-time.

Share this project: