Robotics is increasingly used in the field of healthcare. Oregon Health and Science University is a leader in performing robotic surgery while Japan is focussing on the use of robots for eldercare. However, one application of robotics that draws the attention of researchers in the implementation of this technology in autism therapy.
Autism is a neurological disability characterized by restricted interests and activity and abnormal development in terms of social interaction and communication. The diagnostic symptoms of autism include impairments in nonverbal behaviors like gestures, facial expression, and eye-to-eye gaze, and lack or delay of spoken language. Autism may occur within 3 years of age.
Diagnosis of Autism using Robots
Most of the autism-related problems can be relieved with the help of quantitative, objective social response measurements, which can be implemented through two methods:
- Passive observation of a child while playing or interacting with physicians or caregivers
- Structured interactions with robots that have the ability to produce standardized social “presses” to obtain social responses.
Passive sensors are those used for recording social response information without involving the need for interactions. Social robots’ perceptual systems are capable of functioning as a passive social cue sensor under various circumstances. These sensors can record and analyze data related to the engagement of the subjects in standard clinical evaluations without employing any specific protocol.
Commercial eye-tracking systems have been used for passive sensing where the subjects are required to wear a baseball cap having an inertial tracking system and eyepiece assembly/camera for recording the close-up images of an eye. Different gaze patterns can be observed by autistic adults while looking at naturalistic social scenes.
Interactive Social Cue Measurement
Though passive sensing technologies collect a wide range of information related to social responsibility, the utilization of interactive robots is a distinct approach for evaluating the social responses in detail. The key benefits of this approach include:
- Information regarding low-occurrence social behaviors can be easily obtained
- Ensures standardized, repeatable stimulus and recording methodology
- Quantity and quality of collected data can be increased with the use of simple interactive toys.
Working Example – The AuRoRa Project
The AuRoRA project involves developing a robot in the form of a toy that would serve a therapeutic or educational role for children suffering from autism. The main goal of this project is to enable children with autism to engage in synchronized and coordinated interactions with the environment and thus helping them develop their social interaction and communication skills. The social skills focused by this project are imitation and turn-taking besides communication and interaction skills.
Though children with autism are comfortable playing with toys, the environment is not ideal for their development. Therefore, the robotic platform is an effort to link the gap between the world of learning and human contact and the safe, predictable and stable environment of a simple toy. The robots are used to teach basic social interaction skills to children using imitation games and turn-taking. Moreover, the utilization of robots as objects or mediators will further increase the interaction of autistic children. Pekee, Mel, Kaspar, Robota and AIBO ERS-7 are some of the robots used in the AuRoRA project. The video below demonstrates a fully functioning robot that is capable of sensing via sonar capabilities, sensitive to touch, and a microphone to help communicate with the child. This particular robot could be the next generation tool for helping cognitive development in autistic children.
Example – The Robota
Robota is a doll-like robot whose head, legs and arms are made of plastic components. The main body includes head, legs, arms and five motors having one degree of freedom each in addition to some electronic components. This robot is incorporated with video processing and speech synthesizing technologies via a serial link connection with a personal computer. With the user facing towards the camera, the Robota can imitate the upward movements of the arms of the user by using a motion tracking system. The robot responds to the sense of touch with a small jerk by recognizing passive motion of its limbs using potentiometers. Using speech synthesizing technology, the robot is trained to describe its characteristics and tell its name.
The interaction skills in children with autism can be developed with the help of Robota. It can be used as a tool to evaluate autistic children’s response to social interaction behavior and certain human features like facial expressions and direction of gaze. The imitation and mirroring behavior of the autistic children can be studied using Robota.
Autistic children are a special group of people with large heterogeneity according to an Artificial Intelligence research point of view. Hence it is unlikely to use a specific robotic design. It is necessary to explore the design space of interactive environments and link to the niche space. However, the development of appropriate evaluation and design methods proves to be a great challenge. Various interactive environments and their applications in autism therapy can be assessed and compared upon overcoming this challenge.
Sources and Further Reading
- Dautenhahn.K, Werry.I, University of Hertfordshire, John Benjamins Publishing Company,2004, pp.1-35
- Dautenhahn.K, Billard.A, Universal Access and Assistive Technology, Springer-Verlag (London), pp.179-190.