Digital Horizons: Enhancing Autism Support with Augmented Reality

The findings from the review highlight the diverse technological landscape of AR interventions for individuals with ASD. While a definitive association between specific technologies and outcomes is challenging due to study design variations, some trends and potential insights emerge from the provided data. Concerning the device usage, several studies (Liu et al., 2017; Ahsen et al., 2022; Hashim et al., 2022; Xia et al., 2020; Tang et al., 2019; Sahin et al., 2018; Wang & Lee, 2020; Escobedo & Tentori, 2014; Vullamparthi et al., 2013; Kung-Teck et al., 2019; Bauer et al., 2021; Khowaja et al., 2021; Iftene & Trandabăț, 2018; Lian et al., 2023) primarily employed smartphones or tablets for AR interventions. Positive outcomes encompassed improvements in social interaction skills, communication abilities, shopping skills, and academic performance.

AR technology shows significant potential across diverse applications for individuals with ASD, including cognitive rehabilitation, therapy, emotion recognition, and daily living skills. In educational contexts, AR was utilized to support vocabulary learning, reading, and science education, further demonstrating its broad applicability. Notably, the exploration of AR in fostering technology literacy, though still emerging, highlights the evolving potential of this technology in preparing individuals with ASD for increasingly digitalized environments. These findings emphasize the versatility of AR interventions, supporting both therapeutic and educational outcomes while opening new avenues for future research.

Additionally, smartphones were found to be effective in reducing teachers’ workload and enhancing multitasking abilities. Studies utilizing smart glasses or head-mounted displays (Vahabzadeh et al., 2018a, b; Sahin et al., 2018; Keshav et al., 2017) reported positive outcomes such as reduction in irritability, hyperactivity, and social withdrawal, as well as improvements in social communication, cognition, and motivation. Interventions utilizing Kinect and projectors (Nekar et al., 2022; Wan et al., 2022; Wedyan et al., 2021a, b; Lee, 2021; Asif et al., 2020; Pérez-Fuster et al., 2022; Lakshmiprabha et al., 2014) demonstrated improvements in various areas such as social awareness, cognition, communication, emotional recognition, and response to joint attention skills. AR interventions conducted on laptops or PCs (Chung & Cheng, 2017; Huang & Lee, 2020; Zheng et al., 2017; Mahariba et al., 2016; Vullamparthi et al., 2013; Li et al., 2023) led to improvements in learning effectiveness, social behaviors, concentration, and facial expression recognition. One study (Zheng et al., 2017) utilized an EEG helmet and reported encouraging results in understanding ASD children better. This system provided real-time EEG feedback to assess and adapt training sessions tailored to the individual needs of children with ASD. The results indicated that the adaptive system could improve engagement and focus during training sessions, helping children to better regulate their behaviors and responses. Additionally, the system showed potential for identifying patterns in neural activity associated with specific ASD-related challenges, offering valuable insights for developing personalized intervention strategies.

In line with the findings of Koumpouros (2024), smartphones and tablets dominated the platform choices (61.2%), underscoring their accessibility and diverse functionalities. The shift towards mobile augmented reality has been significant, with mobile devices becoming more powerful and affordable, enabling a wide range of AR applications. With the proliferation of smartphones equipped with advanced AR capabilities, such as depth-sensing cameras and powerful processors, mobile AR applications are increasingly being leveraged to enhance the social and cognitive skills of individuals on the autism spectrum. By overlaying virtual elements onto the real world, these applications provide personalized and interactive experiences tailored to the unique needs of individuals with ASD, facilitating learning, communication, and social interactions. For example, AR-based social skills training apps offer immersive scenarios that help individuals with autism navigate social cues and practice social interactions in controlled environments. Moreover, mobile AR games designed specifically for individuals with ASD promote engagement and sensory integration while fostering creativity and problem-solving skills. Recent research (Berenguer et al., 2020; Lian & Sunar, 2021) underscores the transformative potential of mobile AR in supporting individuals with autism, highlighting its ability to empower and enrich their lives through innovative and accessible interventions. The future of mobile AR for autism holds great promise in providing innovative solutions to support individuals with ASD in various aspects of their lives, promoting independence, communication, and social integration. The future of mobile augmented reality is closely intertwined with advancements in artificial intelligence and emotional computing, which enhances AR experiences.

The notable prevalence of smart glasses and head-mounted displays (18.4%) within the scope of this review corresponds to the timeframe under consideration, reflecting a growing interest in immersive experiences that could potentially benefit interventions targeting social skills and emotional recognition, as indicated in the sub-questions (R1). These wearable devices present a hands-free and immersive medium for delivering customized interventions and support tailored to the specific requirements of individuals on the autism spectrum. With ongoing technological advancements, including lightweight designs, high-resolution displays, and sophisticated sensors, head-mounted displays (HMDs) and smart glasses are increasingly becoming more accessible and practical for individuals with ASD. Particularly noteworthy is the debut of Apple’s latest wearable device, the Vision Pro, which has garnered attention for its advanced features aimed at enhancing accessibility and catering to diverse user needs, including those pertinent to autism. The Vision Pro integrates cutting-edge functionalities such as augmented reality overlays, advanced facial recognition, and real-time environmental awareness, allowing users to seamlessly integrate digital information with their surroundings (www.​apple.​com/​apple-vision-pro/​). However, the high cost of the Vision Pro presents a significant barrier to accessibility, especially when compared to more affordable alternatives such as the Oculus Quest. While the Vision Pro integrates AR functionalities, its focus on mixed reality (MR) and VR applications distinguishes it from more AR-centric devices. Additionally, other innovative products such as Microsoft HoloLens 2 and Google Glass Enterprise Edition 2 offer similar capabilities, presenting opportunities for individuals with autism to enhance social communication, sensory integration, and daily living skills through immersive and interactive experiences. Recent studies (Skjoldborg et al., 2022) have underscored the potential advantages of employing HMDs and smart glasses in the autism community, highlighting their role in fostering independence, engagement, and overall quality of life. Nonetheless, the utilization of HMDs and smart glasses is still in its nascent stage, and researchers are poised to explore the possibilities presented by new and exciting products, including cost-effective alternatives that cater specifically to the needs of individuals with ASD. The affordability of these products inevitably influences their broader adoption (Koumpouros, 2024). Furthermore, the range of other devices utilized, including interactive boards and EEG helmets, demonstrates the exploration of various technological avenues beyond traditional platforms, in line with recommendations to leverage the distinct strengths of different technologies (Koumpouros & Kafazis, 2019). A key point to consider in any application and research is, of course, the potential for autistic people to be hypersensitive to any wearable device.

The prevalence of marker-based techniques, superimposition-based methods, and projection-based approaches underscores the adaptability of AR technology to cater to various needs within the ASD community. Many studies (Liu et al., 2017; Ahsen et al., 2022; Hashim et al., 2022; Nekar et al., 2022; Wan et al., 2022; Wang et al., 2022; Lee et al., 2018; Asif et al., 2020; Chung & Chen, 2017; Sahin et al., 2018; Huang & Lee, 2020; Kurniawan, 2018; Zheng et al., 2017; Wang & Lee, 2020; Keshav et al., 2017; Zheng et al., 2017; Vullamparthi et al., 2013; Alkadhi et al., 2020; Singh et al., 2019; Kung-Teck et al., 2019; Khowaja et al., 2021; Iftene & Trandabăț, 2018; Tang et al., 2019; Cunha et al., 2016; Lian et al., 2023) employed marker-based AR techniques. For instance marker-based AR was used to enhance social communication skills through interactive social scenarios. Superimposition-based AR helped reduce sensory processing issues by overlaying calming visuals onto the environment. Projection-based AR was used in the study by Liu et al. (2017) to support academic skills through interactive educational games projected onto physical surfaces. Positive outcomes included improvements in social skills, communication, learning, attention, engagement, and academic performance. However, some studies (Singh et al., 2019; Lian et al., 2023) reported usability challenges for individuals with ASD when using marker-based AR. AR interventions utilizing superimposition-based methods (Nekar et al., 2022; Vahabzadeh et al., 2018a, b; Sahin et al., 2018; Krishnamurthy et al., 2022; Vahabzadeh et al., 2018a, b; Keshav et al., 2017, 2018; Lorenzo et al., 2019; Pérez-Fuster et al., 2022; Lakshmiprabha et al., 2014; Bauer et al., 2021; Amado et al., 2021; Khowaja et al., 2021; Iftene & Trandabăț, 2018; Tang et al., 2019) demonstrated positive outcomes such as reduction in ADHD-related symptoms, improvements in social communication, cognition, and motivation, as well as high tolerability and usability. Studies employing projection-based AR (Wan et al., 2022; Wedyan et al., 2021a, b; Asif et al., 2020; Pérez-Fuster et al., 2022; Lakshmiprabha et al., 2014) reported improvements in facial expression recognition, social interactions, talking, and emotional recognition. One study (Chen et al., 2016) utilizing overlay-based AR reported improvement in emotional recognition and social skills. Based on the results, marker-based techniques, which involve the use of physical markers or objects to trigger AR content, have been instrumental in creating interactive and engaging experiences for individuals with ASD, facilitating learning and communication in a personalized manner. Superimposition-based methods, where digital information is overlaid onto the real world, offer opportunities for individuals with ASD to enhance their sensory experiences and improve spatial awareness. Projection-based approaches, which project AR content onto physical surfaces, have shown promise in supporting social skills development and promoting interactive play among individuals with ASD. These varied approaches underscore the potential of AR technology to address a wide range of needs within the ASD community, providing tailored solutions for communication, education, and social interaction. Our study is in line with the research conducted by Bouaziz (2020) and Khowaja et al. (2020) highlighting the benefits of marker-based AR in improving social interaction skills and reducing social anxiety among individuals with ASD. Similarly, Porayska-Pomsta et al. (2012) demonstrated the efficacy of superimposition-based AR methods in enhancing communication abilities and promoting social engagement in ASD individuals. Additionally, recent findings from Wedyan et al. (2021a, b) showcased the utility of projection-based AR approaches in facilitating sensory integration and improving attentional focus in individuals with ASD. These studies collectively emphasize the versatility of AR technologies in addressing a wide range of needs within the ASD population, providing promising avenues for personalized interventions and support strategies. Another interesting finding is that iOS (38.8%) has gained more interest in the studied articles, indicating the trend for cross-platform applications. This aligns with the recent findings of Koumpouros (2024) suggesting the development of apps for both operating systems. Additionally, the identification of Unity as the predominant development tool underscores the importance of user-friendly software environments in facilitating the creation and implementation of AR interventions. The necessity for user-friendly software to develop AR applications is crucial, particularly for individuals such as teachers and therapists who may lack programming expertise but play vital roles in supporting individuals with ASD. Koumpouros (2024), Zhang et al. (2022), and Dechsling et al. (2022) collectively emphasize the critical role of user-friendly AR software in democratizing the development process and empowering educators and therapists to harness AR technology effectively in supporting individuals with ASD. Our findings suggest that different technologies used in AR interventions are associated with diverse outcomes in individuals with ASD. While smartphones and tablets were effective in improving various skills and reducing teachers’ workload, smart glasses and head-mounted displays showed promise in reducing symptoms and enhancing social communication and cognition. Additionally, marker-based AR techniques were widely used but posed usability challenges for some individuals with ASD, whereas superimposition-based methods demonstrated high tolerability and usability. Overall, the choice of technology may influence the effectiveness and usability of AR interventions for individuals with ASD, emphasizing the importance of considering specific needs and preferences when designing and implementing such interventions.

The review highlights AR interventions’ focus on communication and social skills, essential for the well-being of individuals with ASD. It also reflects a holistic approach by addressing cognitive function, special education, and therapy to meet diverse needs. We also identified evidence trends across the studies, including common outcomes and intervention strategies, while being mindful of the limitations inherent in the current body of research. Most studies focused on improving social interaction, communication skills, and academic performance in individuals with ASD. However, the diversity in intervention strategies and outcome measures across studies makes it challenging to draw definitive conclusions. Commonalities and differences in findings were discussed to provide a comprehensive overview of the current evidence. For example, while some studies reported significant improvements in social skills, others found no substantial effects, highlighting the need for further research with standardized methodologies. Despite this progress, significant gaps remain in understanding and applying AR interventions for individuals with ASD. The limited representation of attention-related issues, emotion recognition, and rehabilitation training in the literature highlights the need for further research in these areas. Advancements in artificial intelligence and affective computing hold considerable promise for research in emotion recognition (Shu et al., 2018). Additionally, leveraging deep learning techniques in conjunction with EEG signals presents novel opportunities in this domain (Wang et al., 2023; Wu et al., 2024). Thus, it is imperative for researchers to focus on exploring these avenues further. Additionally, the relatively low prevalence of interventions addressing vocabulary learning, daily living skills, ADHD-related symptoms, science vocabulary, and education points to areas where AR technology could be further leveraged to support individuals with ASD. The emphasis on communication and social skills (53.1%) aligns with previous findings underscoring AR’s efficacy in these areas (Yuan & Ip, 2018), while addressing cognitive function, special education, and therapy (8.2% each) highlights the versatility of AR interventions in addressing diverse challenges, consistent with initial hypotheses (R1). Targeting attention-related issues, emotion recognition, and rehabilitation training (2% each) suggests potential benefits for psychological and cognitive rehabilitation, an area warranting further exploration (Bakir et al., 2023). The inclusion of studies addressing vocabulary learning, daily living skills, ADHD-related symptoms, and educational settings (2% each) emphasizes the broad applicability of AR across diverse contexts, as explored in R1. Notably, exploring technology use itself (2%) reflects an emerging interest in utilizing AR to enhance technology literacy in individuals with ASD, potentially opening doors to new learning opportunities. However, the limited focus on specific end-user groups (other than students and educators) indicates a gap in addressing the diverse needs of different populations within the ASD spectrum, as identified in R4.

In the examined articles, most studies targeted children, with varying age ranges and both typical development and ASD diagnoses. Only a limited number of studies targeted students and the final applications were used in school settings for academics and social skills (Vahabzadeh et al., 2018a, b; Keshav et al., 2018; Kung-Teck et al., 2019). Limited studies focused on adults with ASD (Xia et al., 2020; McMahon et al., 2016), while educators and therapists were involved in using or evaluating interventions (Ahsen et al., 2022; Escobedo & Tentori, 2014; Iftene & Trandabăț, 2018) in three studies. Another finding, was the small groups (2–10) of end users involved in the evaluation phase. Larger groups (12–30) were used in studies focusing on feasibility and broader applicability. There was no clear association between number of users and specific outcomes. The review underscores the diverse array of evaluation techniques utilized to gauge the effectiveness of AR interventions for individuals with ASD. The predominance of subjective evaluations (76.9%) reflects the inherent difficulties in implementing standardized assessments within dynamic AR environments (Koumpouros, 2024). Inherent difficulties in using standardized assessments in AR environments include variability in user interactions, differences in AR hardware and software, and the dynamic nature of AR experiences. These challenges can be addressed by developing standardized protocols for AR evaluations, creating adaptable AR content that can be used across different platforms, and conducting longitudinal studies to assess long-term effects. Additionally, the high prevalence of observation techniques (49%) and participant interviews (46.9%) aligns with recommendations favoring qualitative and user-centered evaluation approaches in AR research. However, the limited utilization of validated questionnaires (12.2%) indicates a need for a greater emphasis on standardized measures to ensure the comparability and generalizability of findings across studies, echoing the sentiments of R2. Incorporating more objective assessments (26.9%) can enhance the overall evaluation process by providing a more balanced perspective (R2, R3). Identifying areas for improvement in evaluation methods, the review highlights the relatively low adoption of objective assessment techniques, potentially indicating a bias towards subjective measures (Koumpouros, 2024). Moreover, the infrequent use of control groups in research designs emphasizes the necessity for more rigorous study designs to ascertain the efficacy of AR interventions for individuals with ASD. Aligning with Koumpouros (2024), our study identifies shortcomings in study designs, such as small sample sizes. Looking ahead, researchers should continue exploring innovative evaluation methods and adopting robust study designs to furnish more conclusive evidence regarding the effectiveness of AR interventions for individuals with ASD.

The scarcity of control groups (2%) highlights the need for more rigorous research designs, as identified in R2, R3. Collaborative efforts from researchers, developers, and practitioners are crucial to design and implement well-controlled studies targeting specific needs and utilizing diverse end-user groups, addressing the limitations identified in R4 (Koumpouros, 2021, 2022). Future research directions should consider ethical considerations and real-world implementation challenges (R5). It is important to note that current AR studies mostly demonstrate associations between AR interventions and positive outcomes, rather than causality. Thus, while promising, these findings should be interpreted with caution. Out of the 49 studies reviewed, 38 reported positive outcomes in terms of improved skills or behaviors, besides general satisfaction and social acceptability. As we look to the future, it is essential to address the limitations identified in the review and embrace the potential of AR interventions for individuals with ASD. The limited representation of certain domains and the lack of standardized evaluation methods present clear areas for improvement in future research. Researchers should prioritize addressing these limitations through more comprehensive study designs and rigorous evaluation protocols. Moreover, efforts should be made to ensure the inclusivity and accessibility of AR interventions for individuals with ASD, taking into account factors such as socioeconomic status and cultural background (Pickard & Ingersoll, 2016). Future AR applications could focus on endorsing social skills in a general population in more complex group settings. By leveraging the engaging and interactive nature of AR technology, developers can create innovative solutions to enhance communication and social interactions among individuals of all ages. There is some evidence that AR interventions are associated with the ability of individuals to generalize social and communication skills learned in AR environments to real-world interactions. Collaborative design between researchers, developers, healthcare professionals, and individuals with ASD and their families will be crucial in driving forward innovation and ensuring the meaningful integration of AR technology into the lives of individuals with ASD. By addressing these limitations and embracing the potential of AR interventions, we can continue to advance our understanding and support of individuals with ASD, ultimately enhancing their quality of life and well-being.

Overall, this review paints an encouraging picture of the potential for AR interventions to address various challenges faced by individuals with ASD. The diverse technological landscape, focus on key areas of need, and emerging evaluation methods point towards promising opportunities for further development and implementation. By addressing the identified limitations and pursuing future research directions outlined in R5, the field of AR interventions has the potential to significantly augment the lives of individuals on the autism spectrum.