In the European Project ICAERUS we are working on a Drone Data Analytics Library (DDAL). We envision this library as an open-access repository of the most significant existing and emerging drone data analytics models and algorithms. Analytical models of UAVs are often published behind closed doors, or completely developed in-house. This limits wider adoption, evaluation and further optimisation of the insights these models can provide.
As the UAV domain can be quite broad when it comes to models. We have chosen to focus on a wide array of analytics-implementations of UAV applications. This can range from:
- Photogrammetry
- Statistical models
- UAV spraying simulations
- Machine Learning/Deep Learning Algorithms
- Vegetation Indices calculation
- Fleet management optimisation algorithms and Travelling Salesman Problem.
Additionally datasets are also seen as an essential part in sharing and collaborating of UAV analytical applications.
The beginning of this library is available at the following location:
Furthermore, there are already a few openly available datasets at the ICAERUS zenodo:
These datasets are also tracked in the following GitHub repository:
The GitHub contains the code, models, instructions for the various sub-activities developed in the ICAERUS Use Cases, and ICAERUS partners. However, new contributions related to UAV analytics are welcome, going beyond the ICAERUS group. Some example models contain:
The library is currently being populated from existing work from the five Use Cases that are part of ICAERUS. However, we also welcome outside contributions to this library. Through GitHub anyone can become a member of the ICAERUS organisation and contribute to this library, first steps can be found at the github homepage.
The ICAERUS DDAL is also automatically linked to the ICAERUS Platform. This platform makes it easy to find different models and datasets. The library gives access to cutting-edge models and algorithms. analytics. This can be valuable for enhancing their own research or applications. Participation also opens doors for collaboration with the initiative’s stakeholders. By contributing models, algorithms, and techniques to the library, individuals and organisations can accelerate the overall progress of research and development in the field. Contributing to the library also allows individuals and organisations to share their expertise and knowledge with a broader community. This collaborative approach fosters a culture of knowledge exchange and learning from one another.
Getting started, for users:
Explore the currently included models and datasets in the ICAERUS platform. Every model should contain enough information to get started using the approach.
Perhaps you have improved, adjusted an existing model, perhaps created or changed a dataset and want to share it to a wider audience?
For a dataset:
- Write a short description
- Notify Jurrian Doornbos () to include it within ICAERUS or any support on your activities!
- Structure the files in a logical manner
- Upload to Zenodo
- Make sure there is enough (FAIR principled) documentation
For a model:
- Have a GitHub account, and make sure the model can meaningfully be shared through a GitHub repository
- Follow the ICAERUS repository template or ICAERUS GitHub main page to get started with your own repository for the model.
- Notify Jurrian Doornbos () to include it within ICAERUS or any support on your activities!
