
NAO robotic. Picture credit: Stephen Chin on Flickr [CC BY 2.0]
Created by the French firm ‘Aldebaran Robotics’ in 2008, and purchased by ‘Softbank Robotics Japan’ in 2015, NAO is an autonomous and programmable humanoid robotic that has been efficiently utilized to analysis and improvement functions for youngsters and adults. Greater than 13,000 NAO robots are utilized in greater than 70 international locations around the globe. Just about each lab working in human-robot interplay analysis owns a NAO making it the social robotic that has been essentially the most used within the historical past of the sphere.
In our paper ‘10 Years of Human-NAO Interplay Analysis: A Scoping Overview’ printed in Frontiers in Robotics and AI, we current an outline of the evolution of NAO’s technical capabilities. We additionally current the primary outcomes from a scoping overview of the human-robot interplay analysis literature through which NAO was used.
Technical overview of Nao through the years
Look-wise, NAO hasn’t aged a bit. The infantile anthropomorphic look of the robotic has stayed just about the identical over the past 10 years. Nevertheless, a whole lot of enchancment has been made over the past variations of the robots.

Evolution of the NAO robotic’s technical attribute since 2008
We additionally word that the software program surroundings for the robotic has additionally improved through the years integrating new applied sciences and options as years handed: face recognition and monitoring, semantic engine. The NAOqi API additionally helps numerous programming languages. In 2013, Aldebaran additionally created a developer group. This group platform allowed to share assets, behaviours and featured a discussion board. This has in all probability participated so much within the growth and adoption of robotic with builders and researcher sharing designs of behaviours, and different video games they developed with the robotic.
Quantitative evaluation of the scoping overview

Geographical distribution of the papers included in our scoping overview
The geographical distribution of authors utilizing NAO for his or her analysis in HRI is unprecedented for a social robotic. NAO has been utilized in greater than 50 international locations all around the globe. And we see from the above map that the distribution is nicely break up between North-America, Europe and Asia. However nonetheless, identical to in a whole lot of analysis subject the worldwide north, tends to publish greater than southern international locations.
When it comes to the position of the robotic, researchers have investigated totally different software area with the robotic getting used as a peer or a demonstrator in public areas, an academic and assistant for therapists (for youngsters with ASD, studying incapacity or listening to incapacity) being the dominant ones.

Enter information (left), output information (center), gear (proper) for every software subject
As talked about above, NAO is provided with numerous sensors and actuators permitting for multi-modal interplay. Typically researchers have used the robotic to talk and/or make gestures. On the notion aspect, the robotic’s restricted speech recognition and poor microphone high quality refrained researcher to make use of NLP however as an alternative made them depend on facial, and gesture recognition. We additionally famous in our evaluation that researchers usually used exterior gadgets corresponding to RGBD cameras or tablets to be able to seize different indicators permitting to evaluate the scenario.
Qualitative outcomes
A number of research reported how enjoyable and fulfilling was the looks of the robotic. NAO is definitely not simply an eye-catcher robotic as its portability is very appreciated by the researchers. NAO may be considered a plug-and-play robotic because of its sturdy and simple setup traits. It has additionally been discovered inexpensive in comparison with different humanoid robots.
On the opposite aspect, a number of technical points had been reported corresponding to low battery life and overheating points. The speech recognition limits the interplay distance and context (i.e. noisy environments). Lastly, not a lot embedded autonomous performance might be used within the preliminary variations.
Nao’s communities
Over time, a number of varieties of communities have fashioned across the NAO robotic platform.
1) Academic communities
In 2010, the College of Tokyo piloted a brand new instructional program developed by Aldebaran Robotics that aimed to coach college students to make use of NAO. Together with some reductions to purchase NAO for instructional functions, the corporate launched some ebooks. The books and academic program cowl each technical elements of utilizing the robotic in addition to extra inventive initiatives that may be given as train to college students. The tutorial market definitely helped in making NAO broadly used within the 2010’s.
2) Engineering communities
In 2007, the Robocup determined to make use of the NAO robotic because the official platform for its soccer league. As detailed in the overview, this led to fruitful development on technical capabilities of the robotic corresponding to its locomotion and imaginative and prescient system.
Extra not too long ago, the introduction of the Pepper robotic as a brand new platform for the [email protected] problem definitely can allow us to suppose that it’s going to increase the robotic’s economical life and produce extra advances in social and at dwelling duties with the robotic.
3) Analysis communities
Aldebaran had a robust Analysis & Growth technique in its early days and naturally turned companion of a number of giant European initiatives: ALIZ-E, DREAM, SQUIRREL, L2Tor. These methods had the benefit to type the long run researchers in the usage of their platform and made NAO a dominant robotic in human-robot interplay analysis between 2015 and 2019.
4) Industrial translation

Picture from RobotLab web site visited on December twentieth 2021
Just lately a community of start-ups has woven across the NAO robotic. Offering numerous companies corresponding to instructional workshops, therapeutic periods, or enterprise and advertising companies, the variety of these start-ups have raised these previous few years (e.g. Interactive Robotics, RobotLAB and Avatarion).
Conclusion
In our overview, we current a complete overview on the usage of NAO, which is a exceptional social robotic in lots of cases. To this point, NAO has been uncovered to a difficult but rewarding journey. Its social roles have expanded because of its likeable look and multi-modal capabilities adopted by its health to ship socially vital duties. Nonetheless, there are gaps to be stuffed in view of sustainable and user-focused human-NAO interplay. We hope that our overview can contribute to the sphere of HRI that wants extra reflection and common proof on the usage of the social robots, corresponding to NAO in all kinds of contexts. An implication of the findings exhibits a better want for growing the worth and sensible software of NAO in user-centered research. Future research ought to think about the significance of real-world and unrestricted experiments with NAO and contain different people which may facilitate human-robot interplay.
tags: overview
Wafa Johal
is a Senior Lecturer on the College of Computing & Info Methods, School of Engineering and Info Know-how, College of Melbourne.
Wafa Johal
is a Senior Lecturer on the College of Computing & Info Methods, School of Engineering and Info Know-how, College of Melbourne.
Aida Amirova
a Analysis Assistant on the initiatives of CoWriting Kazakh and Robotic-assisted autism remedy.
Aida Amirova
a Analysis Assistant on the initiatives of CoWriting Kazakh and Robotic-assisted autism remedy.
Nazerke Rakhymbayeva
is a PhD scholar on the Division of Robotics and Mechatronics, College of Engineering and Digital Sciences, Nazarbayev College.
Nazerke Rakhymbayeva
is a PhD scholar on the Division of Robotics and Mechatronics, College of Engineering and Digital Sciences, Nazarbayev College.
Elmira Yadollahi
is a Postdoctoral Analysis Fellow at KTH Royal Institute of Know-how.
Elmira Yadollahi
is a Postdoctoral Analysis Fellow at KTH Royal Institute of Know-how.
Anara Sandygulova
is an Assistant Professor of Robotics at Nazarbayev College.