On Friday, Meta announced PartNR, a new program designed to study human-robot interaction (HRI). This study focuses specifically on how humans and robots work together in home environments. This includes common tasks like cleaning, cooking, and picking up food delivery.
Automatic housework is a dream decades ago, with Rosey being captured most prominently in “Jetson.” The robot maid debuted on Primetime TV over 60 years ago, but remains a popular touchstone when discussing the possibility that sophisticated machines can remove the burden of housework.
But so far, only robot vacuums have made great strides in the market. There are many reasons why other home robots aren’t cracking the mainstream, such as price, reliability and limited features. It certainly isn’t due to a lack of attempts, nor is it due to consumer indifference. This means that no other robots have acquired marks in both cost and feature sets.
Seeing more robots at home could improve collaboration with the people who own them. The first wave of home robots rarely manages chores on their own. Even the vacuum of a good robot sometimes needs help. Meta places PartNR as both a benchmark and a dataset to determine how people and robots work together to get things done in the home.
“Our benchmark consists of 100,000 tasks including household chores such as cooking and cleaning toys,” writes Meta. “We are also releasing a PARTNR dataset consisting of human demonstrations of PARTNR tasks in simulations, which can be used to train embodied AI models.”
Simulation has become an increasingly useful tool in robot deployment, allowing organizations to test in seconds that can take hours to achieve in the real world. However, Meta also stated that it was successful in deploying the PARTNR model outside of simulation. It has already been used in tests in the spot robots of Boston Dynamics. Meta also builds a mixed reality interface designed to provide a visual representation of the robotic decision-making process.
“The possibilities for innovation and development in the field of human-robot collaboration are enormous,” adds Meta. “With PARTNR, we want to rethink robots as our future partners, not just agents, and start research in this exciting field.”
Age Tech holds many possibilities in this category. Labrador’s automated serving cart, for example, provides insight into how to help older people who continue to live independently. However, many advancements in meta varieties aim to address them before such systems gain mainstream acceptance.
Humanoids are another interesting path that has introduced themselves in recent years. Most businesses behind these bipedal robots foresee a future that will ultimately support them at home. That said, pricing needs to drop significantly, and reliability needs to be improved through jumps and boundaries. That’s a large part of why most manufacturers are trying to address the needs of companies first.
With proper scaling and advances in AI, we can imagine a world where humanoid robots deal with general purpose tasks in ways that are useful in both factories and homes. The main stepping stone to that location requires solid progress in human-robot collaboration. Meta, exploring robotics within the wider AI research, hopes that PartNR will help them get there.