For some time now we've been talked to by a friend about an idea for shipment tracking. The use case is straight forward, imagine something like a solar panel or electrical equipment being shipped to places or within a supply chain. The equipment is regularly checked as part of ongoing maintenance and one wants to store a record of those checks including date, time and location. Doing this on paper, while common, is rather unpractical, so a digital solution is preferred. Since we got to meet our friend this spring we took it upon ourselves to implement a prototype of the idea as a surprise. We made it so that it smoothly runs on a phone and played a live demo with two friends, where we create the QR, one scans the QR and one observes the dashboard. Here's a rough idea of what we built including a gap analysis of what would be still needed to make this a viable solution.
| Frontpage of our demo application with three subpages for the three functions we implemented. Since our friends are in South Korea, we made everything in korean. |
At the heart of the idea lies data packaged in QR codes. These codes are created with identifiers for the type of equipment as well as a unique ID to identify each piece of equipment. Once those codes are put on the equipment the fun can begin. Our prototype comes with two functions (three if one counts the QR generation):
- reading a code and sending it together with current timestamp and GPS location to an endpoint
- reading the database and mapping GPS locations of scanned codes on a map.
![]() |
| System sketch including users and what part of the application they interact with. |
With these two basic functions we can illustrate how the solution would be used and how maintenance technicians and operators would report their work to a database.
| A QR code we generated for blue berries. In our live demo, we played one person is the exporter, one is checking/scanning the QR and one is observing the dashboard. |
When showing this to our friend we ran the solution on our laptop and deployed it online using ngrok. A real mvp would obviously be hosted properly. There was no authentication in place, which of course needs to be added. We used a sqlite db with just one table, which should be extended into proper database management and there needs to be ways to add/delete and manipulate the data too. An earlier idea including blockchain was abandoned. Our friend thought perhaps storing the records in a blockchain would make more robust or trsutworthy but given what happened with blockchains over the last few years, trustworthy is not a word anyone would associate it with. (And a normal SQL db is enough anyway).

Keine Kommentare:
Kommentar veröffentlichen