Roadmap
This is an extensive explanation on where we are heading. We have laid out a plan for our road ahead for the further development of the application. Here you will see our future plans and ideas. This roadmap is shared with you so you can see where your contributions are helpful and/ or what you can contribute to.
This is a broad overview of our current roadmap and the direction we are taking. For a more detailed and technical roadmap , please have a look at our Trello board.
If you like the direction in which we are going with this application , please consider supporting us with funding. At the moment we are funding all development made thus far, ourselves because we see the value in this application/ system. The application as it stands today is a reflection of our abilities, drive and passion. Feel free to support us financially, we would love to improve this application and turn it into something that improves the efficiency of many industries, including yours.
2D navigation and mapping
This section is giving some background into how our current two dimensional navigation and mapping system works in order to better explain the three dimensional mapping feature on our road map.
Search for an object or tag and initiate navigation by pressing the find button.
Point the camera of the device towards your containers , so that the device can determine where the device is currently.
If the app cannot find a direct path to the selected container it will display a pop-up directing you to the correct shelf.
Once Tswiri detects a barcode that has a direct path to the container you are looking for it will direct you towards it.
(see next two images)
If the container you are looking for is not on the screen Tswiri will draw an arrow in the direction of the container (Similar to turn-by-turn navigation)
Once the container you're looking for appears on screen Tswiri will highlight the container.
Tswiri will keep an eye out for any containers that might have moved since it's last visit , and keep track of these moves.
3D navigation and mapping
We want to take the 2D mapping we currently have and take it into the 3D space.
To do this we will need to solve PnP equations to extract angle data from the camera feed.
(See figure 1)
3D navigation will eliminate the problem of starting the search on the wrong shelf as the 3D map will be able to guide the user directly to the target.
Accuracy improvements
We are always aiming to improve the accuracy of our mapping and navigation by any means possible, this includes improved filtering, machine learning, and just all round experimenting with different techniques of how to better calculate the position of the device at any time.
Desktop and server applications.
We want to create a desktop and server application in order to enable up scaling of the system. Having the ability to have multiple users that can use their own device to navigate and find containers.
This is important for use cases where a large amount of people need to navigate the same space ex. Warehouses.
Improved object detection
Our goal is to create the best possible object detection system that reduces the amount of labelling that needs to be done by users. The better the system can recognise objects from just pictures the faster you can find your required objects.
At the moment we are considering integration with custom machine learning models , 3rd party image analysis services such as AWS and Google Cloud. We are also looking into Auto ML integration.
Cloud backup
Currently users can export a .zip file and import it again to restore data. We are aiming to provide seamless backup functionality using Google drive and iCloud.
Detection of objects without barcodes/QR codes
Detecting objects that are not in containers and assigning rough co-ordinates to them so they can be located again.
GPS location
Possible application of GPS location when multiple warehouses are used.
IOS
We are using flutter to develop the app so at some point we want to include IOS users.