First round of distillation

[x] Add Dago Cruz Contributions

Open Source AI Artificial Intelligence (AI) is changing and helping in various aspects of our daily lives, from entertainment to complex medicine discovery tasks. The democratization of current AI technologies allows and open-source technologies allow people without in-depth skills in maths, statistics, and programming to create their own AI solutions. Examples of these new open-source platforms with a high-level programming approach are TensorFlow, Keras, and Scikit-learn. However, most of these technologies use specific syntax and grammar and work with a particular programming language. Thus, the learning curve for children can be a complex task. Therefore, research groups have initiatives to design and create open-source platforms to bring children closer to AI. Cognimates is an AI education platform for children 7–14 years old. Tutors can use Cognimates to introduce and teach children on building games, programming robots & training AI models. Based on Scratch programming, children can develop an AI literacy through enjoyable tasks. Machine Learning for Kids (MLK) is another initiative to aware and teach children about AI. MLK is a free tool that uses Scratch and App Inventor to provide hands-on training ML model and build things with them. PictoBlox is another AI education platform for children. PictoBlox allows enthusiastic children to learn AI and ML using pre-build models on recognizing text, faces, and objects; design virtual assistants to make their own projects and solutions. These solutions are examples of how to bring children closer to AI and ML. However, most of these platforms focus on software-based AI solutions. Although they can deploy solutions on real robots (such as JIBO, NAO, others), these robots can be inaccessible for children worldwide, particularly for those who live in poor and developing countries.

Open Source Robotics There is an inherent engagement in children with robots [ref]. Thus, research initiatives have considered this to use robots to enact education solutions. They take advantage of real-time deployments from computers/mobile into real robots based on the Do-It-Yourself approach. Thus, the platforms encourage programming skills in children. Shybo is a robot that combines open-source hardware and software [ref]. Shybo can perceive sounds and react through non-verbal behaviors (movements and lights). The creators claim that Shybo can also be used in educational contexts to support playful learning experiences. Sparky is an Arduino-based mobile robot. Creators provide schematics, 3D model files, and source code underneath are all open source. Using block or code programming, Sparky introduces programming from elementary-age to adults.

bibtex

@article{Lupetti2017,
abstract = {This article presents Shybo: a novel low-anthropomorphic robot for children. The robot, resulted from the combination of open-source hardware and software, is able to perceive sounds and to react through two non-verbal behaviors: hat's movement and lighting. By taking advantage of an open-source machine-learning software, the robot can be easily trained by children. This robot can be employed in research to support human-robot interaction studies with children, for investigating perceptual aspects of robot's features or for investigating children' cognitive abilities. It can also be used for applications in educational context to support playful learning experiences.},
author = {Lupetti, Maria Luce},
doi = {10.1016/j.ohx.2017.08.003},
issn = {24680672},
journal = {HardwareX},
keywords = {3D printing,Child-robot interaction,Machine-learning,Open design,Open hardware},
pages = {50--60},
publisher = {The Author},
title = {{Shybo. An open-source low-anthropomorphic robot for children}},
url = {https://doi.org/10.1016/j.ohx.2017.08.003},
volume = {2},
year = {2017}
}

@article{Druga2019,
abstract = {We observed how 102 children (7-12 years old), from four different countries (U.S.A, Germany, Denmark, and Sweden), imagine smart devices and toys of the future and how they perceive current AI technologies. Children outside of U.S.A were overall more critical of these technologies and less exposed to them. The way children collaborated and communicated while describing their AI perception and expectations were influenced both by their social-economical and cultural background. Children in low and medium SES schools and centers were better are collaborating compared to high SES children, but had a harder time advancing because they had less experience with coding and interacting with these technologies. Children in high SES schools and centers had troubles collaborating initially but displayed a stronger understanding of AI concepts. Based on our initial findings we propose a series of guidelines for designing future hands-on learning activities with smart toys and AI devices for K8 students.},
author = {Druga, Stefania and Vu, Sarah T. and Likhith, Eesh and Qiu, Tammy},
doi = {10.1145/3311890.3311904},
file = {/[doi 10.1145_3311890.3311904] Druga, Stefania\; Vu, Sarah T.\; Likhith, Eesh\; Qiu, Tammy -- [ACM Press FabLearn 2019 - New York, NY, USA (2019.03.09-2019.03.10)] Proceedings of FabLearn 2019 on - FL20.pdf:pdf},
isbn = {9781450362443},
journal = {ACM International Conference Proceeding Series},
keywords = {AI literacy,Child-Agent Interaction,Inclusive education},
pages = {104--111},
title = {{Inclusive AI literacy for kids around the world}},
year = {2019}
}

@article{VanStraten2020,
abstract = {This narrative review aimed to elucidate which robot-related characteristics predict relationship formation between typically-developing children and social robots in terms of closeness and trust. Moreover, we wanted to know to what extent relationship formation can be explained by children's experiential and cognitive states during interaction with a robot. We reviewed 86 journal articles and conference proceedings published between 2000 and 2017. In terms of predictors, robots' responsiveness and role, as well as strategic and emotional interaction between robot and child, increased closeness between the child and the robot. Findings about whether robot features predict children's trust in robots were inconsistent. In terms of children's experiential and cognitive states during interaction with a robot, robot characteristics and interaction styles were associated with two experiential states: engagement and enjoyment/liking. The literature hardly addressed the impact of experiential and cognitive states on closeness and trust. Comparisons of children's interactions with robots, adults, and objects showed that robots are perceived as neither animate nor inanimate, and that they are entities with whom children will likely form social relationships. Younger children experienced more enjoyment, were less sensitive to a robot's interaction style, and were more prone to anthropomorphic tendencies and effects than older children. Tailoring a robot's sex to that of a child mainly appealed to boys.},
author = {van Straten, Caroline L. and Peter, Jochen and K{\"{u}}hne, Rinaldo},
doi = {10.1007/s12369-019-00569-0},
file = {/van Straten, Peter, K{\"{u}}hne - 2020 - Child–Robot Relationship Formation A Narrative Review of Empirical Research.pdf:pdf},
issn = {18754805},
journal = {International Journal of Social Robotics},
keywords = {Artificial intelligence,Automation,Child–robot interaction,Human–robot interaction,Ne

Links

https://thestempedia.com/product/pictoblox/ https://machinelearningforkids.co.uk/ http://arcbotics.com/products/sparki/

air4children / hri2021

First round of distillation #3

bibtex

Links