Socially Assistive Robotics and Discoveries on the Research Path

by Maja J Matarić


  • Ph.D University of Southern California

  • BS in CS at University of Kansas

    • Go Jayhawks
  • Ph.D earned at MIT

  • Center director of CRES, WiSE Chair in Engineering, President of the Faculty, Vice Dean for Research

  • Give much credit to her students

Full bio of speaker

Dr. Maja J Matarić is a professor of Computer Science, Neuroscience, and Pediatrics at the University of Southern California, and founding director of the USC Center for Robotics and Embedded Systems. She received her MS and PhD in Computer Science and Artificial Intelligence from MIT and is fellow of the American Association for the Advancement of Science and of the IEEE, and recipient of the Presidential Awards for Excellence in Science, Mathematics & Engineering Mentoring, the Okawa Foundation, NSF Career, and the MIT TR35 Innovation Awards. She is featured in Michal Apted’s movie “Me & Isaac Newton”, in The New Yorker (“Robots that Care”, J. Groopman, 2009), Popular Science (“The New Face of Autism Therapy”, 2010), IEEE Spectrum (“Caregiver Robots”, 2010), and is one of the LA Times Magazine 2010 Visionaries. Her research into socially assistive robotics is aimed at creating caregiving machines that can provide personalized assistance in convalescence, rehabilitation, education, and eldercare. Her group is developing robot-assisted therapies for children with autism spectrum disorders, stroke and traumatic brain injury survivors, and individuals with Alzheimer’s Disease. Details are found at


  • Enabling technologies

    • Bodies (sensors, effectors… humanoids)
    • Brains (Moore’s Law)
    • Afforability
  • Socio-economic factors

    • Growing aging population

      • In 20 years there will be as many old people as young people.
      • Who will take care of all these old people?
    • Healthcare crisis

    • Increased concerns about safety

    • Tech-savvy youth

  • People and robots can get closer than ever: HRI is finally possible and interesting

Question: What is HRI?

Bridging the Care Gap

We don’t have the people to provide care for the young, ill or aged.

Imagine a robot:

... that can assist a phisical confgitaive therapist/coach
... that is enjoyable to interact with
... that is easy to command and interact with
... that is unobstrusive
... that encourages socialization
... that increases human quality of life
... that can help identify early signs of disorders
... that can provide continiuous support

They call this robot a ‘shepherd / guide’.

A new frontier of Sicence and Engineering

Human centered robots has the potential to benefit both how we do science and how we develop technologies:

  1. Robots as tools for scientific inquiry into human behavior and learning processes
  2. Personalized robot technology helps people, improve health and performance and quality of life

Socially assistive robotics

Robots that help through social rather than physical interaction:

  • Monitoring

    • Early detection
    • decreased risk of injury
  • Coaching/training

  • Motivation

    • Get people to do what they need to do to improve their quality of life.
  • Companionship/socialization

    • Avoid isolation and depression


Robotics enhance, not replace, human care

Research questions

  • Why a robot? The role of embodiment and physical presence.

    • embodied communication
    • body language
    • presence
    • believability vs realism
    • compliance
    • uncanny valley issues! TODO Get a link for Uncanny Vally cause I’ve found this interesting for years.
  • Making friends and influencing people? Social monitoring and steering interaction dynamics

    • Socially appropriate behavior
    • personality
    • engagement
    • influencing human behavior and habits
  • Will it last? Long-term personalized user adaptation.

    • sustained engagement
    • improved human wellbeing
    • Adaption along w/ the user indifinitely


Social time has to be realtime. Chess is not realtime. A conversation with body language has to be done in realtime or you lose engagement with the user.

Autism Spectrum Disorders (ASD)

Those with ASD will live full length lives and are at least as intelligent as the rest of us. How do we make them more productive within society?

  • Children with ASD interact socially with robots in ways they do not with people or computers
  • Robots seem to elicit social behaviors, communication, join attention, turn taking, initiating play, even the first social smile
  • An opportunity to develop robots as tools for ASD diagnosis, intervention and therapy


They look at ways that kids with ASD suddenly begin acting with robots in a way they don’t interact with humans or computers. They do amazing astonishing things they don’t do otherwise.

Stroke Rehabilitation

  • Most stroke sufferers are left with permanent deficits due to a lack of long-term supervised rehabilitation
  • 40% of traumatic brain injury symptoms (TBI) are similar to stroke
  • Rehabilitation requires hours of supervised daily exercise
  • Continual motivation has shown to be a critical aspects of recovery; rehabilitation is depressing.


Stroke sufferers after the 12 weeks of physical therapy often don’t continue working the body to improve


Stroke sufferers will stay engaged but cheat if they can!


Robots are always interpreted as male. Adding a wig and bra to a robot is not cool. Real tests have shown that trying to change the gender of a robot is counterproductive.

Eldercare, Alzheimer’s Disease, and Dementia

  • We’re all headed there
  • Aging-in-place requires a social component to offset isolation
  • Evidence supports physical fitness as effective against Alzheimer’s as medication
  • Research has shown that senior citizens really like robots. Any statements otherwise are not backed up by science
  • Singing games with too-perfect voices (like Frank Sinatra) are counterproductive. This is a good example of Uncanny Valley.

Role modeling: How to be a mentor

  • Be transparent and don’t compartmentalize
  • Expect 100% effort and commitment
  • Demand collaboration. We are a team
  • Be completely honest but fair and mindful of difference. You are not me but this is how it works
  • Be kind but not wimpy. Life strikes again but so what?

Mentoring advice

  • Do what you love and love what you do
  • Don’t take no for an answer.
  • Be true to yourself: speak up, bit wisely (and increasingly with seniority)
  • Do outreach, so you do good and get perspective
  • Refuse the false choice of work vs. family
  • Be excellent and demand excellence of others
  • Form a network of peers; it’s all about people
  • be a mentor and recruit mentors lifelong