Friday, July 6, 2018

Top 5 Advances Needed in Robot-mediated Behavior Intervention Technology in autism and Why We Need Them

The number of children diagnosed and thus impacted by autism continues to rise, and staggeringly so. Recall, that in 1985 children with autism was estimated to be 1 in 2500. Ten years later in 1995, it was 1 in 500. The CDC now estimates (2018) 1 in 59. Associated lifetime costs have risen past $3 million dollars. Medicaid costs for children with autism is reported to be 6 times the normal at over $10,000.00 per year. ABA therapy at $120.00 per hour for 10 hours of therapy per week, puts therapy cost at an annual rate of over $50,000. per year. 80% of children with autism are in divorced families. The human costs over a life time? Special living requirements? Employment potential? Lost productivity? Social lifestyle impacts? How can these be quantified?

To even have a chance at mitigating these impacts early intervention is critical. The efficacy of early interventions are, therefore, even more critical.

The utilization of social robots in autism behavior especially in early intervention is an emerging and powerful force. The past initial phases of academic research clearly demonstrated that ‘robots’ can be an effective resource used in behavior intervention therapies. We now face the challenge of further developing hardware, software and curriculum that improves upon and enables the transfer of robot-mediated behavior intervention technologies to be more broadly used and more cost-effectively deployed to therapists, special education professionals, school teachers and, yes, even to parents, guardians and family caregivers.

Bottom line is that current intervention methods and resources are not sufficient to address the growing need. As technologists working with autism experts we can and must do more with robots.

What is needed?

I helped organize and conduct a just completed workshop held under the auspices of Carnegie Mellon University’s 2018 Robotic Sciences and Systems Conference in June 2018. The workshop was designed to specifically address this very subject. Participants from academia, education and the robotics industry joined-in to contribute their thoughts about the current state and the future design requirements needed to support the advancement of robot-mediated behavior intervention.

Robot-mediated behavior intervention needs to advance to the next level with:

1    1.     More software behavior ‘APPS’ for use with the robots – especially for early age (3-6) use
2.       APPS that are highly customizable to the ‘speed, sound and voice interaction’ needs and capacities of individual children
3.       Robotic platforms that are less disrupted by the environmental conditions of ‘light and noise’
4.       Advanced technological response interfaces for ‘non-verbal’ children
5.       More accurate and more comprehensive data collection progress and performance data capture facility built into the ‘APPS’

The full list is longer but the above five will help significantly move ‘progress and adoption’ measures. The next major technological phase will be machine-based learning algorithms that designs in real time, custom therapy interventions. I will soon address that subject in a future blog.

Michael Radice is Chairman of the Technology Advisory of ChartaCloud | ROBOTTECA and can be reached at