Prof. Kiam seeks an innovator for a planning use-case in open call #2 for innovators. We spoke with her to learn more about this opportunity.
Please briefly introduce yourself
I am an assistant professor at the Institute for Flight Systems of the University of the Bundeswehr Munich. In my line of research, I focus on mission planning for complex aerial operations, with a focus on including unmanned aerial vehicles and on increasing automation in decision-making support systems used for complex mission planning.
The Institute for Flight Systems of the University of the Bundeswehr Munich aims to bridge theory and practice. We work towards deploying algorithms/methods in AI for complex mission planning, and towards increasing usability of AI by investigating human-autonomy teaming. Most importantly, we work closely with system end users, in order to engineer usable AI-enabled systems.
For what specific problem are you seeking a solution?
Disaster-struck areas can become impervious to first responders, causing delay in the rescue operations (e.g. Search-and-Rescue, monitoring of impervious disaster areas) or even danger to the first responders. For that, many first responder organizations leverage unmanned vehicles, and especially unmanned aerial vehicles (UAVs) to explore the disaster area.
To help increase the efficiency of the deployment of UAVs, a viable solution is to encourage manned-unmanned teaming. In this setting, the unmanned vehicles (typically UAVs) will be deployed from another manned vehicle (either a manned aerial vehicle such as a helicopter or a manned ground vehicle). However, this is only possible if the coordination of the unmanned vehicles is highly automatized, as the number of human operators in the manned vehicle is limited. Therefore, an automated mission planning system will be in this setting highly beneficial.
For this purpose, planning algorithms/methods – capable of dealing with temporal constraints, heterogeneous goals and objectives – should find solutions in a heterogeneous state-space with high-level tasks and low-level actions. They need to cope with the dynamic environment, requiring therefore replanning upon changes in the problem instance.
