It wasn’t long ago that humans were dreaming of a new millennium where life would be entirely supported by robots. We thought we’d be flying the family spaceship to work every morning by 2010 or that we’d come home to a house kept clean by a shiny, beeping humanoid on wheels like the Jetsons’.
Though we’re quite there yet, the idea of exploiting technology’s full potential to help humans in their everyday tasks is quickly becoming a reality. Even the film industry has decided to give its utopian tech world a break in order to explore a more realistic relationship between humans and machines. In the meantime, while the arts debate its ethical dilemmas, public and private institutions invest in introducing this kind of technology into real-life situations.
Such is the case with CoCoMaps, a project that aims to implement cognitive map architecture in humanoid robots to encourage complex interactions between robots and humans. The project was funded by the European Coordination Hub for Open Robotics Development (ECHORD++), which bridges the gap between researchers and manufacturers to bring successful experiments and ideas into the market.
A project like this could have interesting ramifications for specific work environments that are highly mechanised and involve little to no human interaction. However, CoCoMaps has different plans altogether. As Þór List,the managing director of CMLabs (the other partner in the CoCoMaps project), explains on the phone, CoCoMaps expect to implement their robots within individuals’ homes in the near future, especially in relation to elder care. “These robots could be sold to elderly people who spend most of their time alone, who perhaps live far away from their families or who get turned down by senior houses,” Þór says. “A system like this would definitely diminish individual costs when it comes to caregiving.”
The concept of machines interacting with people through cognitive maps is not new in itself: Honda released its first functioning ASIMO humanoid robot in the year 2000. However, CoCoMaps aims at a more complex kind of interaction, wherein the robot is required to collaborate with another communicative machine and more than one human simultaneously, thus encouraging social interactions with people while completing different physical tasks. For those unfamiliar with cognitive maps, they can be seen as mental representations used by individuals to learn, store and remember information about their spatial environment.
From a physical point of view, the robots should be able to make out spaces as well as to identify, interact with and manipulate objects. They should be able to stop and change their focus to other tasks or interactions if interruptions occur. In a similar way, from a verbal point of view, the cognitive map supports real-time communication. Although the machine is initially programmed with a series of modular patterns of behaviour, it also learns from the individual and the environment. The cognitive map allows the computer to speak more naturally rather than merely understand simple sentences. Furthermore, it allows for verbal interruptions and two-way interaction, instead of simply providing answers to a stimulus.
Finding multiple, trustworthy caregivers can indeed be an expensive and time-consuming issue. Instead, a humanoid robot with the ability to interact and help humans in their daily tasks could become essential by providing services that a single human caregiver cannot provide at all times. Not only it would allow people to live in their own homes but it would also take note of whether the patient has taken his or her medicines and provide a consistent 24-hour service.
“It would also allow people to obtain or retain their social network without having to impose themselves physically or financially, as the robot is able to help them organise their social life by scheduling calls to family members,” Þór affirms enthusiastically. He adds: “Furthermore, it is able to collect and store information regarding behaviour changes that might be helpful to a doctor, as well as assisting in case of emergency by picking up visual or vocal inputs and forwarding them to a help line.”