Usage Scenario. The material has been omitted pursuant to a request for confidential treatment and the material has been filed separately with the Office of the Secretary of the Securities and Exchange Commission.
Usage Scenario. BannaTree is built on a Django framework. Django is a free and open source high- level Python web framework, which takes care of much of the hassle of web devel- opment. This means that the developer only needs to focus on writing their app without setting up the same framework every time the developer wants to create a web interface. The Django framework follows the model-view-template architec- tural pattern so that different components of a web application are clearly separated. The Django models are subclasses of the Python framework Django. The models contains the fields and behaviors of the stored data. Every single database table is mapped to a model. Figure 7 shows the models which we are using in BannaTree. The Django views provide Python functions which uses web resquests and return a response to a particular templates. The Django templates are used to show all data from the web application. The view functions tell the template what they need to show or how to redirect to another webpage. With this framework the user can add multiple applications in the same project. In the BannaTree application we made use of that function. We created two applications: a mobile web interface for the farmers and a dashboard for the factory and the government. Both applications are hosted on heruko, but can also be hosted on another hosting service which python supports. Heroku is a container-based cloud Platform as a Service (PaaS) which is used to deploy, manage or scale modern apps. On Heroku, developers have the free- dom to focus on their core product without distractions that are needed to host a application. Problems can arise when developers do not have the knowledge about the servers, hardware or the infrastructure of the application. When the application needs to be expanded then only new models or attributes could be added into the app. After implementation of the new models or attributes, Xxxxxx will create new tables for the database. When the new models are built then only the views and templates must be adjusted to be able to show the latest informa- tion on the interfaces. For example, if more information about farms is needed then the application can be expand with more questions about the information which is needed from the farmer. These data can easily be added into the mobile interface as well as the dashboard. Also this BannaTree could be used in multiple countries. At the moment, BannaTree supports only English and Malay, but it is effortl...
Usage Scenario. At the moment, BannaTree provides two interfaces. These interfaces are used by the Agricultural Department, the BannaBee factory, and the banana farmers. Farmers use the mobile interface. They upload every month a report with information about their banana trees. After they filled in the report, they can view these data in tables or graphs views gives them information about if they are on target. This information is also used by the dashboard interface which is used by the Agricultural Department and the BannaBee factory. Both stakeholders have access to the same dashboard which gives an overview of information about which zone a farm is located, which xxxxxx has uploaded their report and different graphs about information about the banana trees and harvested or received bananas. The only difference between the government and the factory is that the factory can add data into the dashboard to get a clear overview of the number of bananas received in kg from the farmers and how many bananas the farmer has harvested from the banana trees. This comparison must be the same in a perfect world, however the discussions with the factory direc- tor and the government showed that there is a gap between the harvested bananas and the usefull bananas to create banana chips.
Usage Scenario. The envisioned usage scenario would be that a 3D scene like ‘The Ranch’ or the ‘50 megatons’ scene serves as input. The scene can then either be manually described or is fed into a machine learning framework. This framework would generate an output like the XML provided below. Technical, semantic as well as contextual descriptions of the scene and the individual assets need to be filled to make them reusable. Otherwise they will not be detectable through a search framework. In a virtual production context the descriptions (metadata) greatly simplify the setup and authoring of an initial scene. Not only can assets be found conveniently, that are explicitly suiting virtual production needs (e.g. real-time capable rigs, meshes with less polygons etc.) but the metadata also serves as guide for the transformation part of frameworks like the search and transformation framework developed within SAUCE. If an asset does not fulfill all requirements of a search, the framework can look up if transformation plugins are capable of automatically transforming the asset to fulfill all search criterias. Based on the labels, the scene could be set up for a usage in the Virtual Production Editing Tools (VPET) developed by FA. Through the labels, movable objects like cars, benches etc. can be separated from static parts like houses. Movable objects are then marked as editable in the VPET system, enabling a user to select and modify it in realtime on set with the VPET tablet clients. In addition the labels can be used to provide scene understanding to e.g. the animation solving engine in order to generate scene aware animations.
