2.2 KiB
layout, title, categories, excerpt, header
| layout | title | categories | excerpt | header | ||
|---|---|---|---|---|---|---|
| single | AI-Fusion: Emergence Detection for Mixed MARL Systems. | acoustic anomaly-detection projects | Bringing together agents can be an inherent safety problem. Building the basis to mix and match. |
|
{: .align-left style="padding:0.1em; width:5em"}
In cooperation with Fraunhofer IKS this project aimed to investigate and detect emergent effects in multi-agent reinforcement learning scenarios, i.e., mixed-vendor autonomous systems (AI fusion).
Emergence in general refers to emerging dynamics of higher complexity (i.e., sum), which are fed by interacting entities (each other and the environment) of a specific complexity level (regarding their policies and capabilities).
{: .align-center style="padding:0.1em; width:30em"}
In this context, we developed a full-stack, high-performance environment in Python, following the gymnasium specification for the training of reinforcement learning algorithms.
{: .align-right style="padding:0.5em; width:10em"}
| GitHub Repo | Read-the-docs | | Install via PyPI | Read the Paper (TBA) |
The main differentiation from already established MARL environments is its ability to employ various scenarios as a combination of modules and configurations. As well as the option to define per-agent observations, including placeholder and combined observation slices (grid-world). Moreover, this environment can handle multi-agent scenarios as well as sequential actions for inter-step observations.
Furthermore, we designed and implemented a Unity demonstrator unit that can load and replay specific pre-recorded scenarios. This way, emerging unwanted and unsafe situations can be replayed and intuitively investigated.