The Trend Monitor has a strong history of providing early insights into emerging social, technological, economic, environmental, and political (STEEP) trends as well as their strategic consequences for the aviation industry. Bauhaus Luftfahrt adopts now a new AI agent-powered approach to develop scenarios for aviation's future between now and in 2035.
The scenario approach has three steps, each combindes inputs of experts with the outputs of proprietary AI agents:
- Identification of STEEP trends for global aviation,
- a uncertainty-impact analysis of the trends,
- and the development of 2035 scenarios based on critical trends.
AI agents are artificial entities founded on a Large Language Model (LLM), such as ChatGPT. LLM-based agents are capable of autonomous decision-making and creativity, task decomposition, using tools, adapting to circumstances and errors as well as interacting with other agents. For the purpose of this project, AI agents can be understood as specialized instances of LLMs. Bauhaus Luftfahrt utilize various multi-agent systems, in which each agent serves a very specific purpose, performs a very specific task.
Take part in our trend monitoring!
As seasoned expert from any field of aviation Bauhaus Luftfahrt invites you to participate in this project. You will be provided the list of trends identified and described by AI agents and asked to suggest trends which may have missed, and also to score the trends on uncertainty and impact - and of course you will also receive the final results and scenarios.
To participate, please send an email to trendmonitor[at]bauhaus-luftfahrt.net
AI agents are artificial entities founded on a Large Language Model (LLM), such as ChatGPT. LLM-based agents are capable of autonomous decision-making and creativity, task decomposition, using tools, adapting to circumstances and errors as well as interacting with other agents. For the purpose of this project, AI agents can be understood as specialized instances of LLMs. Bauhaus Luftfahrt utilize various multi-agent systems, in which each agent serves a very specific purpose, performs a very specific task.
Step 1: AI agents discover influential STEEP trends that can shape the future of global aviation.
Each trend is accompanied by a facts-based description of current observations and available projections, and a citation of the sources that inform the descriptions. These trends are shared experts who give feedback about trends that the AI agents may have missed.
Step 2: Experts score the trends on the key dimensions of uncertainty & impact.
The objective of the uncertainty-impact analysis is to evaluate potential risks or opportunities by examining:
- the level of uncertainty surrounding a trend (regarding its level of development by 2035, its pace of occurrence etc.)
- the potential impacts on different aspects (such as business models, revenue streams, passenger behaviors etc.) and stakeholders, if the trend does occur.
Lessons from developing and applying Bauhaus Luftfahrt AI agents
- Enhance time- and cost-efficiency: AI agents process efficiently and rapidly vast amounts of data from a wide range of sources, at nominal cost.
- Enhance scalability: AI agents to scale to executing several different tasks (e.g. research to brainstorming to creative writing), to include several sources and to examine multiple trend-types.
- Quality through control: Through a combination of agents’ parameter-tuning and well-defined prompts, the agents are forced to focus only on certain manageable tasks, which prevents them from generating hallucinations, assumptions, or baseless predictions.
- Profoundness: Collaborative truth-seeking results in profound results. A complex project like STEEP analysis has been broken down into sub-tasks, which are then assigned to a group of several agents. These agents focus on their tasks autonomously, but subsequently collaborate with each other. This leads to detail-rich outputs that are not possible to obtain with generalized LLMs.
- Relevancy: LLMs only have access to information up until some date in the past. AI agents are given access to up-to-date information.
Bauhaus Luftfahrt develops AI agents through agile software development. This provides a rapid iteration through different agentic-workflow systems, prompts and large language models to identify the combinations that give the best possible results. With time LLMs get smarter and the same applies to the scoring and scenario agents. The result are agents that better identify and describe trends, better score them, and better develop scenarios at a cheaper and faster rate.
References
- Schneider, A. (2024, April 04). Aviation’s metaverse future in four scenarios. TNMT. tnmt.com/metaverse-four-scenarios/
- Wulf, T., Meißner, P., & Stubner, S. (2010). A scenario-based approach to strategic planning: Integrating planning and process perspective of strategy. Leipzig, Germany: HHL Leipzig Graduate School of Management.
- Xi, Z., Chen, W., Guo, X., He, W., Ding, Y., Hong, B., ... & Gui, T. (2025). The rise and potential of large language model based agents: A survey. Science China Information Sciences, 68(2), 121101.