Image credit: Awakening Health
From 2001: A Space Odyssey’s HAL to Big Hero 6’s Baymax to Her’s Samantha, Hollywood finds the idea of a human-robot relationship compelling. The lifelike interactions from these big screen robots presumably stem from advanced artificial intelligence, which give them human-like capabilities and feelings.
In reality, robots haven’t reached that level of sophistication, but robotics companies are pursuing it. They are building robots with integrated AI, computer vision, and machine learning so robots can interact with humans in a more natural manner.
Creating AI-enabled humanoid robots can improve interactions with people and take over some tasks, such as health screening and customer service, and even provide companionship. Human-like robots engage our brains more readily than chatbots or virtual assistants. Our central nervous systems respond well to 3D human-like forms that activate our neural pathways.
The Rise of Robots
Some robots are already “working” in banks, schools, and healthcare facilities. They are effective assistants, but they are not designed to look human.
Hanson Robotics is out to change that. The company created Sophia, a wheeled robot with a human-looking head covered in fake skin called Frubber. Sophia has multiple cameras in her chest and eyes, which enable facial recognition and allow her to interact with people. She can make eye contact, mimic a person’s facial expressions, and gesture with her head, arms, and hands.
Image credit: Sophia/Hanson Robotics
The ability to read and react to another person’s facial expressions is one of the hallmarks of human interaction—and one of the greatest AI challenges. Sophia recognizes human facial expressions using a convolutional neural network designed to mimic the way humans process visual information. The AI model was trained using data from thousands of images of humans displaying seven different facial expressions: happiness, sadness, disgust, surprise, fear, anger, and neutral.
Humanoid Healthcare Assistants
These “social robots” are designed to assist humans at home, in medical facilities, and in retail locations, and their adoption is expected to climb. Globally, the market for humanoid robots will grow from $2.1 billion in 2020 to $7.9 billion by 2025 at a compound annual growth rate of 30.6 percent. Much of this growth will come in the healthcare and elder care markets, as social robots can interact with patients and provide a level of personalization that delivery-type robots cannot.
One example of this is Grace, a humanoid robotic healthcare assistant announced in fall 2021 through a collaborative partnership between Hanson Robotics, SingularityNET, and Awakening Health. Grace is the “younger sister” of Sophia and incorporates much of the same technology. She is brimming with sensors and has embedded cameras, speakers, and a touchscreen tablet. Her head, arms, and legs move independently, and each finger has actuators.
Based on the Sophia 2020 hardware platform, Grace is integrated with SingularityNET on the backend to support advanced AI capabilities. SingularityNET is a decentralized AI marketplace that supports interoperability, allowing developers to create and share specialized AI applications.
Awakening Health, founded by the leaders of Hanson Robotics and SingularityNET, is merging medical care with AI-based technologies to address the needs of an aging population and stressed healthcare workers and systems. Awakening plans to use Grace to provide assistance to medical staff. For example, she can take a patient’s temperature through a thermal camera on her chest and send the data to a doctor using blockchain technology to ensure security. In a time when hospitals are crowded and the elderly are isolated, it will help to have a little Grace.