Breakthrough
Considering robots in the classroom
The important and growing presence of technology in classrooms has given rise to some fear that education technology — a combination of applications, tablets, robots, smartboards, data analytics, and the like — will effectively replace traditional classroom teachers. That said, this particular fear is dwindling, and I’ve found that when you Google “will tech replace teachers?” or something to that effect, it is hard to find a media outlet, blog, political official, or a teacher herself who reports that traditional educators will be replaced in the not-so-distant future by technology that can do do the same thing better, or for less money. Many have written at length to dispel this fear, to advocate for the allowance of technology in many forms to continue to enter classrooms uninhibited by fear. Still there is one hugely important sector of elementary and secondary education currently undergoing what I find to be the most exciting transformation in the history of education technology. I think that this transformation causes the question of whether technology will replace teachers to demand new scrutiny.
“Autistic children may learn better from robots than from human teachers,” reported Judith Burns in a 2012 article for the BBC. This finding was based upon a then-emerging trial run at Topcliffe School, a primary school in Birmingham, England. Head teacher at Topcliffe Ian Lowe explained to Burns that “the robots have no emotion, so autistic children find them less threatening than their teachers and easier to engage with.”
Topcliffe uses a robot called NAO, created by the French robotics company Aldebaran. At Topcliffe, children play games with NAO to enhance both their social and academic skills. Games can be as simple as imitation routines:
NAO raises his hand and the children learning from him raise theirs; NAO touches his toes and the children do the same. NAO can also play vocabulary games with children during which he calls out the name of an animal and a student holds up a corresponding picture. If the student is correct, NAO will whir and cheer, flash its eyes and raise its arm. Alternatively, if the wrong picture is produced, NAO takes on a subdued posture and urges the student gently to try again. NAO does not possess notions such as personal space, so if a child thrusts a photo in its face during a matching game, NAO does not flinch and in turn further confuse and alienate the student as human teacher would likely do in the same situation.
Recently, NAO has also been introduced at The Moody School in Haverhill, Massachusetts as part of Aldebaran’s Autism Solution for Kids Initiative, which the company launched last year out of its office in Boston. NAO has
experienced similar success to Topcliffe in the year that it has spent at the Moody School and the three other schools which have benefited so far from Aldebaran’s initiative.
The articles that describe NAO’s success in special needs classrooms are overwhelmingly positive, understandably so; NAO is responsible for leaps in social and academic progress among students at Topcliffe Primary and at the Moody School. Its success rate is reportedly astounding. But I cannot help but remark upon the difference in tone between these articles and those that describe the advent of ed tech to traditional classrooms, which generally carry a more sober, questioning tone. More than anything, I am curious about how the schools that have already benefitted from NAO have introduced the tool to parents and teachers, whether there has been pushback, and for what reasons. If the use of the robots in special needs classrooms is truly to become mainstream — and due to the reportedly astounding success rate, I hope that it does — then I believe it will be hugely important for the early adopters to publicize as best as possible the entirety of their experience so that those who come later are not blindsided by its challenges.
If the use of the robots in special needs classrooms is truly to become mainstream then I believe it will be hugely important for the early adopters to publicize as best as possible the entirety of their experience so that those who come later are not blindsided by its challenges.
Although the first, NAO is not the only robot currently in use to educate children with autism spectrum disorders. In 2013, Robokind Robotics partnered with the University of Texas and the Dallas Autism Treatment Center to create Milo, a child-like robot which has been used not only in the treatment of autism but in diagnosis as well. Traditionally, diagnosis of child autism takes place through social interaction and speech exercises, and therefore typically cannot happen until a child is able to speak. This is problematic because the sooner a child is diagnosed, the more effective his treatment. Thus Milo, with its ability to interact with children through consistent and measured body movements and facial expressions, is able to succeed at early autism diagnosis where humans cannot. And Milo is set apart from even NAO in its ability to make ‘facial expressions,’ which arguably lends Milo even more potential to prepare children with autism for social interactions.
Romibo is a third in the ranks of educational robots. Unlike NAO and Milo, it is not at all anthropomorphic, rather, it resembles a small, fuzzy mound with antenna. Still, Romibo has a screen that displays expressive eyes, and it is able to track eye contact. Developed by Origami Robotics at a National Science Foundation research and engineering laboratory, Romibo was designed to be affordable. While NAO and Milo cost $5,000 and $7,990, respectively, Romibo is currently $698 and its creators are working to drive its price down to between $200 and $300. Romibo has not been quite as extensively documented in mainstream media as NAO and Milo, however, the Autism Daily Newscast reported in 2013 that “Romibo plays several games, including I-Spy, Tell me a Story, and Simon Says,” which are designed to teach children basic academic and social skills.
“Something very special happens when you put a machine and person together,” Maja Mataric told Robotics Business Review. Mataric says that she would like to see robots not only in classrooms but on playgrounds as well. “Imagine if you had these robots who are buddies of kids with autism. I bet that all the kids will be fascinated, and suddenly you have a very natural social link, and the robot can be teaching all the kids things.” However, she cautions that it is not exactly clear why robots are so effective in the education of children with autism spectrum disorders because autism itself is so complex.
Robots have begun to do what teachers simply cannot: connect successfully and consistently with students who have autism spectrum disorders. While tablets, robots, and online classrooms may attempt to aid human teachers in mainstream classrooms when budgets are cut, or when a class is too large for a teacher to manage independently, technology is coming to be viewed as an essential component of the education of students with autism. It seems predictable to me that, when put in the context of the fear that human teachers will become obsolete in the face of technology, this success would be threatening. It seems strange that there is very little public dialogue regarding concern for special education teachers becoming obsolete, or other, broader concerns about the use of robots in the education of children with autism. spectrum disorders.
Though any public concern for the replacement of teachers in mainstream classrooms is largely misguided, I think that the concern is positive in that it forces scrutiny of the way in which technology is integrated into classrooms, and at what cost. Resistance to technology’s introduction into classrooms in response to the fear that teachers will play a less prominent role — although I believe it to be most often excessive — helps to keep education technology honest and staid in its application. Conversely, I find the apparent lack of criticism of tablets and robotic technology in special needs classrooms to be potentially worrisome. The breakthroughs that the integration of robots like NAO and Zeno have made are truly that: breakthroughs. Unlike in more traditional classrooms where technology has been helpful in important but smaller ways — augmenting test scores by a few points, aiding teachers in the management of large classes — the use of technology in special needs classrooms has been revolutionary, and it is hard to criticize the mechanisms of unprecedented results.
Resistance to technology’s introduction into classrooms in response to the fear that teachers will play a less prominent role helps to keep education technology honest and staid in its application.
Still, these breakthroughs are relatively recent. Students at Topcliffe primary and the Moody School have known NAO for only a year or two, not to mention that NAO and robots like it are currently so expensive that it may be some time before they are used widely in public education, so it is likely that the excitement simply has yet to subside and, when it does, I’m hopeful that reevaluation and skeptical public dialogue will follow. It is also important to remember that special needs classrooms do not stand on their own — they are most often part of a larger educational system and thus do already factor into public dialogue more generally speaking.
Moreover, in many cases of children with autism spectrum disorders, the ultimate goal is to integrate the child into a mainstream classroom when possible, as stated by the headmaster of Topcliffe. This illustrates the fact that
human teachers remain the ideal, even as robots and iPads outdo human teachers in the breakthroughs required to equip autistic children with the education they require. As Carl Clement, an Aldebaran Robotics developer who has worked on NAO, said, “as they feel more confident, they move on to the next person, who this time may be real.”
Zoe Wood is a senior at Columbia University. This piece was originally published here.