Trustworthy AI
Our data analysis is based on trusted artificial intelligence. The development was funded as part of the Digitalization Focus Program with support from the Austria Wirtschaftsservice Gesellschaft mbH (aws)
Artificial intelligence (AI) refers to the field that focuses on equipping technical systems with human-like abilities: logical reasoning, learning, planning, and creativity. The key hardware components in these systems include computers, sensors, and actuators.
In terms of software, AI relies on machine learning and deep learning-based programs. The core idea of machine learning is that a computer program automatically improves its performance through newly acquired experiences (data). Results and predictions are continuously optimized by feeding evaluation feedback back into the program.
Machine-Learning | Neuronal networks
Behind machine learning are so-called neural networks.
Neural networks enable the classification of information from data—in our case, image data—based on multiple criteria and similarities. The applied criteria are grouped into labels (e.g., “traffic light” or “car,” or in our case: “vital tree” and “weakened tree”). The data is then categorized accordingly, and the results are presented in a suitable format.
Deep Learning
Deep learning refers to a computer’s ability to learn from raw data instead of relying on human-programmed algorithms, allowing it to independently develop rules for classification.
The European Commission emphasizes the development and use of “trustworthy AI,” which is characterized by three fundamental principles—both from a technical and societal perspective: legality, ethics, and robustness.
Even with good intentions, AI-based automated decision systems can cause unintended harm. Therefore, their development and evaluation require particularly high standards of transparency, accountability, privacy, fairness, and reliability.
Trustworthy AI
It must be ensured that the development, implementation, and use of AI systems meet the following requirements for trustworthy AI:
- Primacy of human action and oversight
- Technical robustness and security
- Privacy protection and data quality management
- Transparency
- Diversity, non-discrimination, and fairness
- Societal and environmental well-being
- Accountability
Sustainability
Society – Economy – Ecology
We share a common goal of sustainable forestry with our customers. Therefore, it is a given for us that both the development and use of FESTMETER’s solutions for classifying coniferous trees align with sustainability guidelines and trustworthy artificial intelligence principles.
It is essential for us to ensure key attributes such as safety, reliability, robustness, transparency, and traceability in our AI-based forest monitoring.
Technical robustness and security
Technical robustness and security are key pillars of trustworthy AI. Therefore, we rely on proven software and hardware as the foundation for our systems.
Whenever possible, we also utilize or integrate existing data and systems to ensure that users can continue working in their familiar environment, such as web mapping within their GIS application.
Privacy Protection and Data Quality Management
Both data and the privacy of all stakeholders are protected throughout the entire process by state-of-the-art security measures.
Transparency
Our goal is to ensure transparency, explainability, and traceability throughout the process chain, thereby strengthening our customers’ trust in AI.
Clear and understandable information and visualizations help us convey our concept of trustworthy AI-based data analysis, further reinforcing confidence in our solutions.
Diversity, Non-Discrimination, and Fairness
Due to the nature of our application (plant physiology – assessing the vitality of trees), there is no risk of human discrimination through our AI. However, we also acknowledge and uphold our social responsibility in the development, distribution, and application of our technology.
Societal and Environmental Well-Being
In this context, we see our role as twofold: strengthening trust in artificial intelligence through our trustworthy AI concept and promoting societal and environmental well-being by implementing sustainable and eco-friendly solutions in our processes.
This includes energy efficiency and data minimization, ensuring that our approach aligns with both technological responsibility and environmental sustainability.
Accountability
With the trust our customers place in us by sharing their data, we also take on the responsibility of accountability. Our priority is to ensure verifiability of our content and processes, providing customers with a dedicated contact person and an online platform to submit inquiries or complaints.
Additionally, consent for data usage is clearly outlined in our Terms and Conditions (AGB) and detailed in our Privacy Policy.