New Modular Architecture Improves Home Robots' Ability to Assist with Complex TasksCornell Researchers Develop MOSAIC System for Collaborative Cooking with Household Robots

Household robots have the potential to assist people with a variety of mundane chores and tasks, from doing the dishes and laundry to cooking, cleaning, and organizing. Despite recent efforts by roboticists and computer scientists to enhance the capabilities of home robots, many of these machines still struggle with more complex and creative tasks, such as cooking in collaboration with humans.

However, a team of researchers from Cornell University has developed a new system called MOSAIC that aims to bridge this gap. This modular architecture enables robots to perform intricate household tasks that require close interaction with humans, such as interactive cooking. The system was described in a paper published on the arXiv preprint server and tested through a series of real-world experiments, demonstrating its potential to assist humans in cooking various recipes.

The MOSAIC system is composed of different modules that handle various aspects of the task at hand. This includes an interactive task planner, an architecture for object recognition and robot movement planning, and a model that predicts human movements. By leveraging pre-trained models for general tasks like language and image recognition, and specialized modules for task-specific control, MOSAIC is able to efficiently collaborate with humans.

To evaluate the system's performance, the researchers conducted 60 experimental trials using two different robotic systems, the mobile manipulator Stretch Robot RE1 and the tabletop manipulator Franka Emilka Research 3. These robots worked closely with a human user to prepare six relatively simple recipes, including salads, soups, and a tuna sandwich. In addition to these cooking trials, the individual modules of MOSAIC were also tested in picking, human motion forecasting, and online evaluations of the task planner. The results showed that MOSAIC was able to successfully complete 68.3% of the collaborative cooking trials with a subtask completion rate of 91.6%.

While the system showed promising results, the researchers note that there is still room for improvement. For instance, the system has not yet been tested on more advanced subtasks such as cutting, rolling, and spreading, and its performance may vary in different kitchen environments. The team plans to address these limitations in future research, with the hopes of inspiring further advancements in assistive robots for household environments.

Steven Russell
Steven Russell is a proficient entity from the Technology field. He completed Master’s Degree in Computer Science and Technology. He was engaged in the formation and administration of computational systems in his previous firm. He is associated with Industry News USA from last 2 years. Due to his command over the technology field, he has become the head of the Technology section in very less time period. “Latest gadgets” is the thing that attracts Steven the most.