Paper PDF,  POSTER,  SLIDES,  VIDEO, A3DE-HRI WORKSHOP WEBSITE, ARXIV

Abstract— Previous work has observed how Neurodivergence is often harmfully pathologized in Human-Computer Interaction (HCI) and Human-Robot interaction (HRI) research [66, 96 , 102, 105]. We conduct a review of autism robot reviews and find the dominant research direction is Autistic peoples’ second to lowest (24 of 25) research priority [24 ]: interventions and treatments purporting to ‘help’ neurodivergent individuals to conform to neurotypical social norms, become better behaved, improve social and emotional skills, and otherwise ‘fix’ us– rarely prioritizing the internal experiences [83] that might lead to such differences. Furthermore, a growing body of evidence indicates many of the most popular current approaches risk inflicting lasting trauma and damage on Autistic people. We draw on the principles and findings of the latest Autism research, Feminist HRI [103] and Robotics [48 , 49] to imagine a role reversal, analyze the implications, then conclude with actionable guidance on Autistic-led scientific methods and research directions.

Cite bibtex format:

@inproceedings{hundt2024autismrobots,

author    = {Andrew Hundt and Gabrielle Ohlson and Pieter Wolfert and Lux Miranda and Sophia Zhu and Katie Winkle},

title     = {Love, Joy, and Autism Robots: A Metareview and Provocatype},

booktitle = {Assistive Applications, Accessibility, and Disability Ethics (A3DE) Workshop at the HRI Conference},

year      = {2024},

eprint    = {2403.05098},

archivePrefix={arXiv},

primaryClass={cs.RO},

url       = {https://arxiv.org/abs/2403.05098},

note       = {Website: \url{https://is.gd/autismrobots}}

}

Cite ACM Reference Format:

Andrew Hundt, Gabrielle Ohlson, Pieter Wolfert, Lux Miranda, Sophia Zhu, and Katie Winkle. 2024. Love, Joy, and Autism Robots: A Metareview and Provocatype. In Assistive Applications, Accessibility, and Disability Ethics (A3DE) workshop at the HRI Conference. ArXiV: https://arxiv.org/abs/2403.05098 Website: is.gd/autismrobots 

We consider 12 reviews related to Autism robots since 2020:

Exceptionally robust systematic Autism robot reviews are (Wallbridge et al. 2024), (Salimi et al. 2021), and (Kewalramani et al. 2023). Wallbridge et al is the strongest review, coproduced with autistic people and the absolute strongest review; covering the process of familiarizing Autistic people with robotic systems while noting withdrawals from studies and areas of limited evidence. Salimi et al. [85] observe significant faults in prevailing Autism research methods, stating that, “The main limitations of current studies [on Autism robots are the] shortage of [Randomly Controlled Trials (RCTs)], low power, and bias.”

No review except Wallbridge et al seriously considers the research priorities Autistic people set for themselves (Sec. 2.3). Some reviews accurately note the lack of high-quality efficacy evidence for these devices, but more inaccurately overclaim efficacy. Three use inclusive language (Sec. 2.2.1), but all others use harmful deficit language and generally tend to promote research well-known to risk harm to Autistic people.

We Cover Current Autism Research:

2.2.1 Language and Feedback. Avoid ableist language for positive, scientifically-backed research communication, which is Autistic priority 5 of 25 (Sec. 2.3). Onboard uncomfortable feedback, and avoid derailing it.

2.2.2 Autistic Strengths. Research should actively seek out autistic strengths.

2.2.3 Double Empathy Problem. Autistic-to-Autistic and neurotypical- to-neurotypical communication are each typically effective, but neurotypical-to-Autistic communication frequently involves miscommunication Non-Autistic people, e.g. researchers, clinicians, and parents, tend to demonstrate a lack of reciprocity [41, 43], or symmetrical exchange as equals where neither Autistic nor non-Autistic people have a dominant position.

2.2.4 Triple Empathy Problem—Negative Communication Impacts on Autistic People in Healthcare Settings.

2.2.5 Autistic Camouflaging (Masking). Autistic people tend to seek genuine connections with others, but in adverse or unsafe environments, many Autistic people must hide Autisic traits to appear ‘more neurotypical’—a process known as camouflaging or masking. Masking is not necessarily a choice [72] and has known negative mental health impacts.

2.2.6 Diversity of Autistic People and Support Needs. Co-ocurring disabilties, race gender, etc.

2.2.7 Autistic Joy. Autistic Joy.examples are stimmig, joyful information sharing, stimming, special interests, and more.

Full details are in the Paper PDF.

Copyright  2024 Andrew Hundt, Gabrielle Ohlson, Pieter Wolfert, Lux Miranda, Sophia Zhu, Katie Winkle
The pastel neurodiversity symbol was made by MissLunaRose12 File:Pastel Neurodiversity Symbol.png - Wikimedia Commons.