Rice and IBM Explore Watson-powered Robot Designed to Aid Elderly and Caregivers

IBM Research and Rice University have collaborated to create the prototype IBM Multi-Purpose Eldercare Robot Assistant (IBM MERA), a first-of-its-kind Watson-enabled application designed to aid the elderly and assist caregivers. Watson is a question-answering computer system developed by IBM.

According to the United Nations, the number of people aged 60 years or older is projected to grow by 56 percent by 2030. To help improve eldercare resulting from this rapidly growing demographic, IBM has opened a new “Aging in Place” environment in its ThinkLab in Austin designed to mimic the types of interactions elderly people have in their homes. By leveraging IBM MERA, the Internet of Things and other cognitive-powered technologies, IBM can study how data from atmospheric, motion, audio and olfactory sensors could be used by the ecosystem of caregivers to potentially improve health care and wellness as physical or environmental conditions change.

“Now is the time to invest in, care for, protect and empower our aging population so they can live more independent lives,” said Arvind Krishna, senior vice president of IBM Research. “Our new research on ‘embodied cognition,’ which combines real-time data generated by sensors with cognitive computing, will explore how to provide clinicians and caregivers with insights that could help them make better care decisions for their patients.”

IBM created the prototype robot with students and faculty from Rice’s departments of Electrical and Computer Engineering and Psychology. The robot is being hosted inside the IBM “Aging in Place” research environment. IBM MERA will be used to perform several tasks: help study innovative ways of measuring an individual’s vital signs, such as heart rate, heart rate variability and respiratory rate; answer basic health-related questions; and determine whether an individual has fallen by reading the results of an accelerometer.

IBM MERA runs on the IBM Cloud and a Softbank Pepper robot interface and uses IBM Watson technologies and CameraVitals, a technology co-developed in Rice’s Scalable Health Labs by Ashutosh Sabharwal, professor of electrical and computer engineering, and Ashok Veeraraghavan, assistant professor of electrical and computer engineering. These labs specialize in developing continuous, passive and automatic sensors to simultaneously measure bio and behavioral markers such as vital signs and blood perfusion.

CameraVitals calculates vital signs by recording video of a person’s face.

These technologies allow IBM MERA to obtain fast, noninvasive readings on a patient’s heart and breathing measurements that can be done multiple times per day. Combined with IBM Watson Speech to Text and Text to Speech applications, the camera can also view whether a fall has occurred and provide information for caregivers.

IBM MERA is also designed to interact with individuals using IBM Watson Speech to Text, Text to Speech and Natural Language Classifier applications so researchers can study how they could receive answers to health-related questions, such as “What are the symptoms of anxiety?” and “What is my heart rate?”

“The IBM Multi-Purpose Eldercare Robot Assistant represents the powerful impact that results when leaders in academia and private industry bring their best to bear on pressing societal issues,” said Rice Provost Marie Lynn Miranda. “We are delighted to work with IBM on this critical project, which provided an opportunity for our students and faculty to collaborate with IBM’s best Age and Ability researchers at the IBM Research Lab in Austin.”

In addition to Sabharwal and Veeraraghavan, the main collaborators at Rice are Margaret Beier, associate professor of psychology, and Christopher Fagundes, assistant professor psychology.