HRI is a vastly wide research area which express the components of the interactionprocess between the humans and the robots. HRI deeply addresses how the behaviors,intelligence, perceptions of the robots towards human impact in the social adaptabilityand acceptability of robots. So its essential to know what the research ndings in HRI inthe context of humans behaviors and attitude towards the robots, in order to understandsuch components in HDI environment.
The perceptions and attitude shown by the human towards robots is a very importantfactor. More specically how the reaction of humans having a robot as a companion is acore point of HRI. To assess the humans attitudes and perceptions towards the robots astudy was conducted by Dautenhahn (2005) with 28 participants in a practical settingwith use of supplement questionnaires as well. The results showed that the participantsliked the idea of having a robot as a companion or assistant and few of them preferred tohave a robot as their friend. The most important nding of this research was the humanlike behavior or human like appearance was not an important factor to accept the robotas a companion. Some other interesting results that could be nd was the test unitswere recommended that robot to be highly controllable and they didnt like it when therobots come too near to them.
That means still there is a potential gap between fullyautonomous robot and humans. These traits are very important to be considered whenimplementing a social companion drone.20Humans behavior towards the robots can be impacted by how a robot is approachingthe human. A study was done by Dautenhahn (2006) to investigated how users prefera robot to approach them while the humans in a sitting position.
The purpose ofthis research was to nd a path planning system for robots. 38 participants in a non-laboratory setting and 15 participants in a controlled setting were contributed in thiscase study. Feelings of comfort amongst the test units were measured while allowing therobots to approach them in dierent directions, distances and speeds. The results haveshown that the test units did not like the approach of robot from the front and it makesthem feel uncomfortable or even like threatening. And it showed that most test unitspreferred the approach of the robot from either the right or left hand side, It doesnt givesan uncomfortable feeling as it approaching from the front. Specically most test unitspreferred the approach from the right side.
So this provides some insights about theBackground Literature Review 14approaching direction, essential distance between drone and person and the essentialspeed of robot in order to result a successive interaction between robot and human.These concepts also give an idea about how we can utilize these research ndings in thefuture drone companions as well.192.2.2 Human Drone InteractionEven though HDI is an emerging and a newer research area a number of researches havedone with the purpose of evaluating the interaction between drone and humans. HDIcan be mentioned as another sub area of HRI since already the UAVs have approachedthe current trend in world of research forming ying robots which is a novel categoryof robots.
As we discussed in the previous section there were so many researches donein the eld of HRI to evaluate the interaction between robots and humans. Up to now,in the eld of HDI also a considerable amount of researches have addressed the humansbehaviors towards drones. Some of those research nding will be discussed in this sectionA case study was conducted by Christ et. al.
in 2016 in order to investigate about thedierent levels of autonomy of drones and how it can aect the user experience (UX).The study was conducted as a student competition and the students were advised tobuild four drones with dierent levels of autonomy. The evaluation of the perceivedUX was done through 24 semi structured interviews. Two types of autonomy for thedrone were implemented, full autonomy that make all decisions by itself, and semi-autonomous that become autonomous as an alternative when the operator wants it to17. The results of case study showed that the participants liked both fully autonomousand semi-autonomous systems and they enjoyed interacting with them regardless of thelevel of autonomy, but still the semi-autonomous drone was mentioned as the mostlyenjoyed drone type among them 17.
In another recent research study by Cauchard in 2016 describes how emotional traits canbe encoded to the ight paths of drones. The study was performed to investigate whetherthree dierent ight patterns which represents emotional states and personality traitscould be recognized by the participants. In this study a concept of natural HDI and howthe natural interaction can be supported by adding an emotional component is discussed.The researchers hope that it would also increase the possibility to create intelligentinteraction between the human and drone. Parallel workshops were conducted to nd outBackground Literature Review 15which personality traits would be expressed best by the drone.
By using questionnairesthey discovered how recognizable the emotions would be. The participants were rstasked to observe the drone ying pattern for each ve tasks which represent personality.Then they were asked try to guess the emotion represented by drones behaviour. Theresults showed that it is possible to the participants to identify how the drone behavesto represent a certain emotion. The most interesting nding of this study was theencoding process of emotions to a ying drone, and how the users were able to perceivethose dierent emotions observing their ying patterns. This gives us a idea in how toutilize the drone with dierent interesting approaches in order to make it more social18.
Another Wizard-of-Oz elicitation study was conducted by Cauchard et. al. in 2015 toinvestigate how naturally the human interacting with drones using gestural techniques.
According to Cauchard There are many advantages of developing drones as autonomouscompanions. Moreover he mentions that an autonomous drones will be vital in the casessuch as outdoor traveling or jogging, since it would be hard to use a remote controlwhile the user conducts other activities. The study results showed that participants feltextremely comfortable interacting with the drones. Some participants were resembledeven as they are with a pet or another humans, by calling the drones by name etc. Theuser study was done as a test session and interaction tasks need to be executed by theparticipants were written in the task cards. The results showed that the main interactiontechnique used by the participants was gestural communication, even though the voicecommands was also used by few participants, many of them felt discomfort when talkingwith drones.
But after getting more familiar with the drone, many of them tended touse voice commands. This gives some other useful insight for the implementation ofcompanion drone with dierent capabilities which can attract the humans perception.21