Imagine a team of robots working together seamlessly in your home, each understanding its own strengths and weaknesses to tackle complex chores. This isn't science fiction, but the promise of a new system called EMOS, which uses the power of large language models (LLMs) to revolutionize how robots collaborate. Currently, coordinating multiple robots, especially those with different designs and abilities (like drones, wheeled robots, and legged robots), is a huge challenge. Traditional methods require painstaking manual programming and struggle to adapt to new situations. EMOS changes the game by allowing robots to generate their own 'resumes,' detailing their capabilities based on their physical design. This lets them understand their own limitations and the strengths of their teammates, much like humans do in a team setting. For example, a drone might realize it's the best for finding a lost object on a high shelf, while a wheeled robot with an arm is ideal for picking it up. EMOS doesn't just stop at resumes. It uses a hierarchical system where the robots first have a 'group discussion' to plan the task, dividing it into smaller parts and assigning themselves roles based on their resumes. Then, they work in parallel, each robot focusing on its assigned subtask. Researchers tested EMOS on a new benchmark called Habitat-MAS, featuring a variety of challenging household tasks in simulated multi-floor homes. The results were impressive. EMOS consistently outperformed other methods, demonstrating the importance of self-awareness and collaboration in robotics. While EMOS currently operates in simulation, it paves the way for truly autonomous robot teams in the real world. Imagine robots tidying up your house, preparing meals, or even assisting in complex tasks like construction or disaster relief. The future of robotics is collaborative, and EMOS is a major step towards that future. However, challenges remain. Scaling the system to larger teams and more dynamic real-world environments requires further research. Improving the 'common sense' reasoning of LLMs to prevent errors in task planning is also crucial. Despite these challenges, EMOS offers a glimpse into a future where robots work together as effectively as humans, making our lives easier and more efficient.