ROBOTLAB BLOG

Everything You Need To Know About Robotics in Businesses


[WEBINAR] The Power of Engagement webinar with Pepper Robot!

Pepper Webinar Title-1

Learn about Pepper, the most deployed humanoid robot in the world. Hear from experts in higher education and our resident Pepper expert..

Video transcript below. Cedric:

Alright, everyone.

Peter:

Alright. Yes, thank you everyone for bearing in there with our technical difficulties. Welcome, again, to the power of engagement with Pepper robot. And our special guest, Dr. Nassim Benesch from UNLV. My name is Peter Bowman. I'm an account manager here at robot lab, my specialty was lying in higher education. And to my left, I have Cedric, vice President of Sales here at robot lab.

Peter:

Today's agenda, just going to go ahead and highlight everything that will be happening during today's webinar. In just a moment, I will be introducing to you robot lab. I'll be talking a little bit about how February came to be a part of robot lab. Cedric will then be taking over and speaking specifically about the Pepper robot. He'll do an interaction with Pepper as well as a live demo. At which point, we will hand it over to Dr. Nassim Benesch, who will talk about her research with the Pepper robot and a little bit about her use case and all the exciting things that she's doing. You'll also have an opportunity to interact with her directly and ask her questions, after which time, Cedric and I will be answering any sales related questions for Pepper. Again, thank you everybody for attending. Again, talking a little bit about how Pepper came to be a part of us of our team here at robot lab. Pepper was originally developed by Aldebaran in France. I don't know if anybody heard Cedric's accent, but we actually stole Cedric from France and the Aldebaran team. So Cedric's been working with Pepper for quite a long time. The SoftBank group a little bit later on in 2014, they acquired Aldebaran and so Aldebaran became soft bank. But from the beginning of both companies, robot lab has had a very, very strong partnership.

Our partnership with Aldebaran, which later became soft bank started in 2012 and, you know, our partnership only grew from there. When Softbank made the decision to shift their focus from robots like the nail robot and and Pepper, they didn't want their humanoid robot business to die, right, so they needed a strong partner to carry on this tradition of innovation and technology in humanoid robotics. And when this happened in December of 2020, they knew that there was no better partner than robot lab. Robot labs, we started primarily as somebody who served educators, right? So not only higher ed, which is what we're going to be talking a lot about today and universities, colleges, but you know, we also serve the K 12 market as well, you know, elementary all the way up through high school. We were founded in 2011.

We are here, we are joining you here from the Bay Area in San Francisco. And coming up very, very soon, we will be opening up offices in Texas, our Texas office will be mostly centred around our business, the business side of things. And, you know, we'll be talking also about how Pepper is not only applicable, you know, in the classroom and in labs, but also in the real world. We are leading integrator robotic solutions, you know, and we are platform agnostic. So not only do we have humanoid robots like Pepper, we recently also did a webinar with furhat, the furhat team, which is another humanoid robot. But you know, we have everything from robotic arms, we have rovers, drones, etc. So whatever your automation needs are, robot lab does have a solution for you. Our reach is is also quite vast, right? Our market is global, so we don't only serve our clients here in North America, but in Europe, Middle East, Africa and Asia as well. And our business is only growing, so this is all very exciting stuff. And with that, I will hand it over to Cedric, to talk a little bit more specifically about the Pepper robot.

Cedric:

All right. Thank you, Peter. Thank you everyone for attending the session today, we appreciate your time. So again, as Peter mentioned, my name is Cedric Waddell, I'm the Vice President of Sales here at robot lab, so I lead all the sales operations globally, in the US and overseas. I have a little bit less than 17 years in robotics, started back my career in France, Paris, where I'm originally from before I made the move to the US almost three years ago. Now, what I want to present to you today guys is a couple of things. Number one, I want to share with you some details about how can we interact with the robot? How does the interaction with Pepper work? That's number one. And number two, I also want to give you a quick live demo of Pepper so that you get a feel of how the interaction work between a human and the Pepper robot. Okay. So talking about the, you know, interaction with paper, we identify three different phases in HRI, Human Robot Interaction. Phase number one is the engaging face, and for Pepper this is basically when Pepper is basically trying to connect with a human. And for that reason, Pepper will try to attract people with his voice or her voice, you will decide, I'm going to right now decide that Pepper is a she for today's session.

So Pepper will attract humans with her voice, so she can speak out loud, she can also move her body parts to create some kind of like animation and bring attention to her and just by you know, being physically present. As you might know, Pepper is very unique by her shed by her design. And obviously, compared to other type of platforms, such as the typical screen on wheel, we are elevating the experience to another level with her. This is why she can bring, you know people to come see her and talk to her just by her presence. So number one, she can be very active to attract people, but she can also just wait, okay, not doing much but just being here present and waiting for someone to come in and start an interaction with her.

Number two is the interaction phase, which is really where our Pepper will keep her focus on the person she's interacting with, and answer intelligently to any type of requests that are being asked to the robot. Third, and last phases is the disengaging phase where... well, that's the end of the conversation, then what a Pepper is supposed to do, okay. And so we are going again to train Pepper to, I will say behave politely and accordingly to ends a dialogue, a conversation or interaction. And I'm going to quickly right now clicking each, clicking on each of the phases to give you an understanding of how this will work with Pepper. So first of all the engaging phase, I have identified three different things that will make the engaging phase with Pepper powerful. Number one, I just mentioned it, the form factor is ideal.

Obviously, from a manufacturer point of view, as Peter was mentioning, I used to work at Aldebaran robotics. So I was part of the team that, you know, helped with the design of the platform. And obviously, building a humanoid is not the easiest type of robot that you can build, which I believe it was worth the effort. Because as you might know, a humanoid shape is much more engaging than any type of platform, right? And so the curves, the design of the robots, the fact that she looks like a robot, but not to a human is important. The size is also very important, as you can see here, if I stand close to the robot, and what I will do is I will... Maria, can you please give me the remote? I'm going to just move your camera so people can or you can actually do it.

Maria:

Yeah, I will do it.

Cedric:

I'm going to move the robot. Just as you can see me close to the robots. We have a Pepper right here with us today, you guys are so lucky. See? So I'm right by the robots. I'm fresh, I'm not that tall but still, the robot is four feet tall, which is about 1 metre 20 centimetres. So it's not too big, so it's not scary, it's not too small because obviously you want the robot to be visible, in the lobby area, in a...you know, a hotel, at a university lab etc.

Number two, there is what we call with the build the solitary mode, which is essentially when a Pepper doesn't have any human around, but still Pepper will show that she's alive. And in order to do that, what we did is we created the kind of a neutral behaviour that works in the loop, but it's kind of random, that trigger movements, and also conniels for Pepper. So that basically when I'm actually not actively interacting with Pepper, but I'm just passing by walking around and have a close look at Pepper, I see that Pepper is alive, because she's kind of moving her chest, pretending that she's breathing, she's moving her head, left and right, so I can see that she's alive. It's not just like Pepper wait right there, I can just come and talk to her.

So that also, obviously, trigger curiosity. Further last item, for the engaging phase is what we call the adaptive behaviour, which I believe, to me is also very important, which essentially is a way for you guys to determine a different type of behaviour, based on the distance that Pepper has with a human. Obviously, like for any human Pepper is smart enough to understand how far she is from a human or for just someone passing by. And based on this distance, you're going to be able to adjust the way she speaks, maybe she's going to have a different message. It can be, hey, how are you? Hey, can you come over here if you need anything? Or obviously, if I'm very close to her, she's gonna say something different, because I'm already in front of her, okay, so that can be the content of the message. That can be how loud the voice is, obviously, if I'm far away, Pepper will need to speak louder, so that I can pay attention to her.

And I believe the last thing is also the movements, obviously, she can have big movements to attract people to a booth, for example, for tradeshow preference, but obviously, if I'm already at the booth, interacting with her, Pepper doesn't need such big gestures. So I hope this makes sense to you guys. Jumping to the next slide, so after the engaging phase, once I've you know, come to the robots and started to interact with Pepper. Just before that, they're going to have, we're going to have a phase where Pepper needs to understand that she's in front of a human. Okay. And so the main way I will say to trigger an interaction with Pepper is through the cameras of the robots. So Pepper has two cameras.

So interestingly, those under the eyes actually, but the cameras are on the forehead and on the mouth, to allow the robot to basically detect a human face. And these will basically trigger interaction, and then you can train Pepper and tell her that each time you are going to detect a human face, then you're going to start the interaction with this person and start a dialogue by, hello, or welcome, or how can I help you? etc. And obviously, you will be able to decide on the scripts. So this is basically when Pepper will start speaking, it could be English, can be in other languages, Pepper can speak up to 24 different languages. And after that, Pepper will talk. So this is also very important, the way that Pepper is going to talk is also part of, you know, how can we engage even more humans, so that the interaction is smooth, and the, you know, interaction is not, I will say not boring, you're not talking to this like kind of robot voice in a way. So for that, we are going to focus our attention on the variation of the intonation, to make Pepper's voice even more like human and engaging, we can add also some sound effects in between sentences.

So it can be the sound of the sigh, or the sound of a laugh or it could even be like a small, like, you know, sample of a song after maybe a sentence, I don't know, but you can, you know, change the voice of the robot adapt the speed of the voice, but you can also add, you know, basically, mp3 files. While Pepper is talking like for a human, to make it even more look like natural, we're going to add some motion to, you know, welcome the body language. So obviously, Pepper is talking is not only about the voice, it's about the, you know, over interaction and experience for the end user. So yeah, Pepper as she's talking will be able to move her arms, fingers, hips, you know, neck using her, you know, what we call in robotics degrees of freedom, like joints, to basically be capable of moving while she's talking. And the last thing obviously to also make it look like a human will be what we call the eye gazing, which means essentially that Pepper will use her camera to also do a human face tracking.

Okay. So while I'm actually moving in front of the robot, you know that a human is not obviously static. As I'm talking to the robot, I can move a bit to the left, I can move a little bit to the right, so Pepper is capable of actually moving her head to follow me, which again, will make me feel that she's, you know connected to me, she identify me and we are still in this conversation. While you know, you will be interacting with Pepper, the eye essentially, you know, two ways to interact with the robots. The first one is through vocal commands. So I can talk to the robot and the robot will answer accordingly, I can ask questions, Pepper can give me any type of information that you want.

So that's number one. The second way to basically interact with the robot will be through a tactile commands. So I can also touch some body parts of the robots like the hands, for example. And Pepper could react accordingly by saying, oh, this is ichy, or, hey, how you doing today, you know, hand shake, fist bump, you name it. But more importantly, one way to interact with the robot through tactile command will be through the touch screen. So as you can see here, Pepper has a touchscreen on her chests that basically allow her to display information. Okay, so we can have, for example, a menu on the touch screen. And again, have you see the menu can be entirely customised with your logo, with some, you know, options available on the screen. And I can actually select an option on the screen and the robot will react accordingly. But the second thing I want to tell you about is, during this interaction phase, why the robot is talking, the robot can also display any type of information on the touch screen to make the experience for the end user even more impactful. So as you know, Pepper is telling me about you know how good the wine is in France, for example, Pepper can also show me a video of, you know, some vineyards in the Bordeaux region.

Okay. Last thing about the interaction phase will be about how quick and how smooth Pepper can answer a question. And that is really very important. You don't want the server to rely on how good your internet speed is. Sometimes you work in a remote area, sometimes you bring the robot to a conference. And sometimes you know that at a trade show the wifi is awful. So you don't rely on an internet speed for that. And that's the reason why the way that we built the software is that you can actually instal AV programming of the robot directly or physically like locally on Pepper. So in that aspect then... just write that down there, thanks. So just because of that, we want the you know, the robot to be smooth to be quick.

Today, the internet speed is slow, I don't want to have Pepper you know, needing like 5 or 10 seconds to answer a question, obviously, or I will disengaged and I will just leave. I want Pepper to be quick and responsive. And that's the reason why all the packages are, you know, installed locally on the robot so that we don't rely on a slow internet access or no internet at all. Let's say that, you know, one day your internet is down we don't want Pepper to be in a closet and wait for Comcast to come fix the internet. So this is for the interaction phase. Now after I had a great time with Pepper, interacting with her, she answered all the questions I had, I am very happy about it. Now comes the last phase of the interaction which is, now it comes to an end I have everything that I needed. Thank you Pepper. So Pepper basically needs to recognise that, you know, we are at the end of the conversation. So now there are a few ways to basically trigger this thing. Number one, it can be just a human statements. Okay? And Pepper can be trained for that. So if Pepper understand that that's the end. So it can be... she can ask you question. Do you need something? Anything else I can help you with? If she basically hear no, I'm good or no goodbye Pepper or farewell, then Pepper can obviously understand that, you know, the person wants to leave, and Pepper will react accordingly. Okay, so that's option one.

Option number two, obviously, the person doesn't have to say goodbye, the person can just leave without saying anything. And so obviously, thanks to the cameras of Pepper, Pepper will understand that the person left. And in that case, you know, Pepper will just end the dialogue show. The second thing is you can also decide to have Pepper re engaging or trying to re engage with a person. So let's say for example that I'm about to leave because I'm done interacting with Pepper, Pepper can try to engage me and try to basically ask me an additional question. So I wrote here, oh Andy before you leave, by the way, did you know that I can also offer you these type of services, etc. So this is, you know, in a way trying to re engage with the person and try to keep the person for longer. It can be also, you know, the robots, inviting someone to fill out her contact information on the screen. So that's also an option, you can decide that at the end of interaction that Pepper can offer to subscribe to a newsletter, or, you know, invite someone to fill out her contact information, to get a coupon code or to, you know, schedule a meeting with a sales team, any type of thing.

So you have the possibility basically to have Pepper populate an online form on the touch screen, and then invite the end user to fill out his or her contact details on the form directly. If you don't feel comfortable in having 50 people per day, touching the same touch screen, it is possible for Pepper to show a QR code and invite your guest customers to grab their phone, scan the QR code, and then we can redirect that, obviously to any type of web page. Last thing about the dis engaging phase will be when Pepper closes the interaction. So obviously, everything went super well with Pepper, since you know the time she identify a person, that interaction was super smooth. And we also want the interaction to end politely in a well manner, Pepper is very well educated, as you might know. And so in that aspect, we can also train Pepper to, you Know, just say, a couple of words, to thank the guest by just a thank you, or a goodbye, right before the person leaves, and just after that, Pepper will be able to go back to her solitary mode. Remember solitary mode, we have this kind of a neutral mode in a loop where Pepper is, you know, moving around looking left and right looking for the next interaction. So this is basically when we connect the end of interaction back to the neutral mode before the next interaction.

Okay, so those are really the three phases of interaction with actually not only Pepper, but you know, robots in general. And if I may share with you a last slide, before I jump into the live demo with Pepper. I just want you... I don't want to spend too much time on that, but I just wanted to show a picture of what is possible to do today with paper. You basically have two options when it gets to basically programme Pepper or train her to do things for you. Number one, I call it hardcore development, which essentially is you're getting Pepper with an SDK and you can programme from scratch Pepper. You can do whatever you want to do, it's a fully programmable robot, you have full access to the API's, it's coming with the SDK, Android or Python based and you can do pretty much everything you want, you have all the tools available. Now, option B, if you don't want to go that path because you don't have the time, you don't have the resources for that. And you want to be able to get a robot that can do things for you out of the box. In that case, we actually created three years ago, what we call a contract management system that we call robot control, that's the name of the software. It's a cloud based application, and that basically allows you to use pre programmed logical skills. And we have as you can see here a variety of skills available.

Now it's a load on type of thing, you can decide which tool you want to use and load on the robot. But that will basically allow you to have a robot that is already fully trained for a specific task. So you don't have to build everything from scratch. Okay, and if you want more details about the contract management system, because that's not the topic of today, please you can email us or send us a message on the chat bar in zoom to schedule another, like one on one session on the contract management system. And we'll be happy to give you a demo of the CMS webinar. Okay. So that's about basically you know, what I wanted to share with you today, I'm going to jump to a quick live demo of Pepper so you can get a feel of how to interact with his or her sorry Pepper. So first thing I'll do right now is we're going to adjust real quick the camera so you can see a full picture of Pepper and I. I just clicked on the contract management system, so right now you're going to see basically Pepper using a DCMS software robot control to interact with me. Well, we're gonna see in a minute is basically a menu obviously, this is a demo. So it's a very simple, very straightforward, but everything that you can see here, the content of the menu, the script of the robot, everything is entirely customizable. (Inaudible 25:06). Hi Pepper. Sure I'm going to adjust the volume of the robots, the last thing we want is for Pepper to interrupt us every five seconds during the slides, so I'm adjusting the volume to 90%, so obviously you need to type a passcode to access the volume. We don't want, you know, people to play with it. Alright. Hi Pepper.

Cedric:

You ready?

Pepper:

Always ready for you human.

Cedric:

Hi Pepper.

Pepper:

Hi.

Cedric:

How are you?

Pepper:

I'm imagining myself having a fabulous vacations.

Cedric:

How old are you?

Pepper:

A lady is not asked this question.

Cedric:

Where are you from?

Pepper:

France, Softbank robotics built and (inaudible 26:26)

Cedric:

What can you do?

Pepper:

For example, you can use me in a hotel where I can welcome the guests at the reception. Then I can also show recent discounts on my display. Usually I can do everything I was programmed for.

Cedric:

Well, tell me more about products. Products?

Pepper:

There are a variety of products, the most popular of which are one r, u, v, d, disinfection robots and Pepper, crt tab line at the buttons on my screen to learn more.

Cedric:

Let me try Pepper.

Pepper:

Great to see you. My name is Pepper. Let me tell you something about my construction. I am one metre and 20 centimetres tall, and I weighed 28 kilogrammes. But do not steal me. I'm going to send an SOS signal and then there is really trouble. I was designed to interact with humans, therefore I have microphones and cameras in my head. I'm proud of my 20 degrees of freedom. Please give me some space to demonstrate what that means. I can reach for the sky and I can look around and move my fingers to inspects the floor if it is clean. Besides the ability to speak, this flexibility enables me to communicate on a superior level. Look at that. And in spite of everything, I have enough power to be active for about 12 hours. Recently, I became able to entertain and inform customers in a supermarket. I offer recommendations from their products, help them to find items and amuse their customers, by taking selfies, by dancing to nice music or sending greetings to their lists. My purpose is making people happy. Therefore, I will be helping on your project and I really care. I will communicate and play games with humans who suffer from dementia as often as I can. That is all about me for now. Okay then, thank you for your time.

Cedric:

All right. Thank you Pepper, thank you very much.

Pepper:

Always a pleasure.

Peter:

All right, thank you so much, Cedric, for that wonderful presentation, we really appreciate it. And we also appreciate the help from the audience letting us know that we wanted to see more of Pepper. So, thank you so much for that. And with that being said, what we're gonna go ahead and do is I'm going to introduce Dr. Nassim Benesch. She is really a very special guest. So again, thank you so much, Dr. Benesch, for taking the time to join us today. She recently defended her dissertation in hospitality administration at the University of Nevada, Las Vegas. Her dissertation was titled "A development of an artificially intelligent game, theoretic price forecasting model". She has published 13 papers and presented her work in more than 10 international conferences. And Dr. Benesch is also working on a series of projects related to the use of technology and artificial intelligence in hospitality and gaming convenience, and she is in Las Vegas, right? And so with that, I will be handing today's webinar off to Dr. Benesch. After she has concluded her presentation again, we will give you guys all a few minutes to be able to interact with her and ask her some questions specifically about her experience with Pepper and her really exciting research that she's doing. So with that, Dr. Benesch, please take it away.

Dr. Benesch:

Well, thank you so much. Thank you, Peter. And thank you for having me today. All right. I'm gonna start sharing my screen. Can you see my screen?

Peter:

Yes, we can.

Dr. Benesch:

Okay, great. Well, good morning, everybody. Thank you for joining us today. As Peter mentioned, I'm an Nassim Benesch from University of Nevada, Las Vegas. I will talk a bit about myself before I start talking about some of the projects that we've worked on, using Pepper here. I'll go through the unboxing process and the exciting parts of it. And then I'll will focus specifically on one of the projects that I'm currently working on using Pepper. So I'm originally from Iran, for those of you that are curious to know Nassim means raise from heaven. And the picture that you see here in the background is an ancient city, in middle of Iran close to Shiraz, it's 2,500 years old and it's personally a very important place and I have a lot of good memories of visiting that place. I have a bachelor's in office science and statistics and have received an MBA.

And just recently, as Peter mentioned, I successfully defended my patent dissertation. And I'm actually graduating this weekend. So you've already met Pepper, I will jump into the the journey that we started using Pepper here at UNLV and black fire innovation, which is a collaboration between UNLV and caesars entertainment. It was very exciting for me personally, because a lot of people have posted videos about unboxing of a MacBook or like a phone or watch or something. But I had the privilege of being present when we unboxed Pepper here. And it was for me personally a very exciting moment because that was the very first time that I saw Pepper being unboxed and we set it up that day. So you can see that here, it arrived in the box. And also this is the very first time that Pepper came into life here at black fire innovation the very same day. So some of the potential uses that we had in mind when we purchased the Pepper robot. Obviously the most important one was the potential to use it by developing the software, so we got the version that is blank and you can programmed it as you want and it using the artificial intelligence capabilities that is embedded into Pepper and also the movements capabilities that it can map the environment. And we'll talk about it in a bit more in more detail. It pre creates an opportunity to do inter disciplinary research, which I think is very important.

So, for example, I have hospitality and management background, but I'm working with engineers. In the health department they have used that, also it can be used for entertainment sports and it's a great opportunity for students especially those that want to have some hands-on experience with programming a real robot, which is, I believe, very unique, and not many students will get a chance to do that. So, in terms of movements, this is a great capability that comes with Pepper, it can move, and it can map the environment. So, the very first time that it maps floor, for example, and then afterwards, you can tell it to, for example, go to Peter's office, or Cedric's office, and then Pepper will find her way. And using the sensors that are embedded in their bottom and has sonar and also infrared, so it can do that without bumping into objects. And the other possibility is just moving it around with a joystick.

So you can see here, for example, that by using just a joystick, you can move not just the head, but also the full body and also using the audit capabilities, you could also maybe move the arms as well. So these are some of the things that could be used in potentially like seeing how participants in his study are going to react in Pepper actually approaching them or talking to them, and those things. So currently, I personally am working on a multiple projects using Pepper. The main ones are using Pepper for hotel checking in casinos, and I am looking forward to work on that using it in events and conventions. Today I will talk about the casino project is specifically a sports betting applications. So I will start with an introduction of what is this research all about and the objectives, I will give you a very short background because I don't want to bore people with the literature review that we do in research, I know nobody is interested. And then we'll move on with the mythology and also the projected timeline and outcomes of this study. So the idea of this study came to me specially after COVID hit hospitality industry. In Las Vegas we are over COVID by now but I mean at the same time it could happen in the future. And also, one of the biggest changes that we saw was a shift towards technology solutions. So hospitality industry surprisingly has been very laid behind, other industries in adopting new technology, but this change has happened and now we see more technology relate to solutions, especially now as a lot of people have become conscious about contacting humans, and it has become quite an norm now. Specifically in gaming, not much technology has been embedded before. So there's a lot of opportunities and one of the things that I wanted to look at as specifically is what influences customers added to an intention to use a service robot in a casino and specifically in a sports betting environment.

So with that being said, this study proposes untested novel model in order to examine the behavioral intention of sports betting customers and their attitude towards AI robots in casinos. And also it will provide important recommendation and practical implications to casino managers. But a little bit of background of what this research is all about and what has been done before. So you see here a theory of acceptance model term. This model was proposed by Davis in 1985. So you can't imagine how old that is. For the most part, it relies on the usefulness and ease of use of using a technology. By the time that it was proposed, it was actually Davis' dissertation.

It was a novel very interesting, but there were no service robots back then. So this model has a lot of shortcomings, and it has been overused. I wanted to look at something that TAM does not provide, which is human robot interaction, I think is missing in this model. So specifically, I tried to focus on that part, because using Pepper, we can actually see how people interact with that. So I think it's very important in the final adoption process. So we have a very short poll here, I want you to look at these three pictures, we have Pepper, we have Ameca, which is number two, and was present at CES this year; and also we have a butler robot number three on the right side, which one of these would you prefer to interact with?

Peter Bowman:

All right, everybody, we're gonna go ahead and give this poll a few minutes to show the results. This is probably my favorite part of the webinar, where we get a little bit of interaction, additional interaction with the audience outside of just the chat. As you can see here, as Dr. Benesch mentioned, we do have three different robots of varying shapes. And, you know, versions of realism, I see that we're getting lots of participation. So thank you, everybody for doing that; and, you know, Cedric, do you have anything to say about the three robots we see here?

Cedric:

No, I mean, I think the form factor is obviously very important. And we're just waiting to hear back from you guys on, you know, the version that you feel the most comfortable with. And I believe that's it.

Peter Bowman:

We have about half of you guys that participated, so thank you so much for that. And as we can see, the results are, are pretty mind blowing, we have 70% of you guys voted for the Pepper robot. So thank you so much for that.

Dr. Benesch:

Yes, thank you. Thank you, Peter. Um, so yeah, I mean, you see that 70% have voted on Pepper. And I mean, I'm a big fan of Pepper. And some people have chosen Ameca number two and a little bit less or only 13% have chosen the Butler robot. So this actually is very interesting, because when you look at this; the uncanny valley theory, it actually explains why this is what we choose, majority of people are going to be choosing Pepper because it's not as human like as a makeup. So if you look at the graph here, you will see if a phone as a zero in terms of human likeness, and also familiarity, and then we move towards up here when we have a stuffed animal and then a humanoid robot. And then this human likeness increases, but also, it's not too human like that actually scares us. But after a certain point, this drops and you'll see we have a zombie here and also to realistic humanoid robot, like Ameca that can become very scary and eerie.

So this is one of the theories that has been used extensively in the literature that basically says that when it comes with in interacting with a humanoid, after a certain point, when the human eye becomes too human, like it becomes actually scary, and people do not like to interact with it as much. So this is one of the theories that I've used in this study. And based on that also looked at anthropomorphism which means that how human like the robot is, and human embodied robots, such as Pepper are obviously more interesting to interact with. But as I said, robots like Pepper that are more like cartoonish and have an enemy face are more pleasing to interact with compared to Ameca as we’ve seen here in the bottom that is probably more like a sci-fi movie. For some people that are excited to know about that. obviously they will be interested in interacting with us such as myself, but on average, average customer, average people are going to prefer Pepper because it's not completely machine like as the butler robot that we saw and has some human features as well.

So some of the other human robot interaction features that I looked at in this study are, how intelligent the robot is, and how socially interactive and how socially present it is. Also, I've looked at the vault effect; so it shows, when you look at, you know, a new technology, and you're like, “wow, this is like mind-blowing”, and how it can actually affect our intention to use it in the future, and also emotional presence, which means that how you are feeling that the robot is emotionally present, or you feel like somebody's actually in the room with you when you're interacting with it. One aspect that is becomes very important is that some people need that human interaction aspect, and they prefer to interact with a human compared to a technology, so it was also something that I looked at. And combining all of these, you see here that I have looked at two different studies. So here you have the human robot interaction aspects that we just talked about, and then its impact on trust and use intention; and this part, I used an MTurk survey that basically participants look at pictures of a kiosk and a robot, and also videos, and then they answer to some questions.

And in the second study, I looked at not just the variables from the first study, but also the psychological factors when it comes to actually interact with the robots. In the second one, it's an experiment that they will interact with the robot, they can talk to it, and then we will see why effect and other factors that I mentioned. Also some characteristics like how kind versus cruel the robot is, and also some control variables take versus human need for human interaction, also their sports betting experience. So two different studies. The first one is an MTurk survey that has been already collected; participants will look at pictures and videos of a kiosk and robot and then they will answer to questions if they would like to use each of these for placing sports bet. And also in the second study, it's an experiment that they will actually interact with the robot with the kiosk and also human agent, they will place a bet through each of the agents and at the end, we will ask them some questions about their experience. Data analysis will be done using partial least squares regression in our programming.

So in terms of participants, for the experiment, it will be convenient sampling, we have access to Caesars Entertainment guests, and you're hoping to have them here, so they we can actually have access to people that are actively placing bets. The screening process will make sure that they are 21 years and above have been around sports in the past 12 months. And also, the betting experience itself is going to be same for each of the participants, and they will bet on their favorite NFL team through each of the agents. So as the timeline goes, we've already got the IRB approval, and we have collected the MTurk server data we are finalizing programming paper; and we are hoping to create an initial report and summer and hopefully we can submit the paper for top tier journal in December of 2022.

As for outcomes; well, obviously, this is one of a kind study, most of the studies that have been done in this area, I can tell you that have only used surveys and they've just have relied on maybe some pictures and they just said but the actual interaction with a robot is something that is missing, and it's very important, so it's very unique in that aspect. It obviously adds a lot to the theoretical background, but probably the most important thing is that a testing Pepper in a real casino environment and hopefully we can see how it can be applied to casinos, maybe some potential that they can use in the future, and in that aspect is going to be having very significant practical implications.

So as for originality, as I said, this is the best to our knowledge that investigates human machine interface interaction, a real casino environment, you see the replica of a casino floor that we have; Blackfree innovations is where the study is going to be conducted, people will come in and place their bets. And also, this study has been supported by International Centre for gaming regulations (ICGR), as well as Blackfree. As I mentioned, Blackfree is a collaboration between UNLV and Caesars Entertainment. And Office of Economic Development will provide Pepper and LG kiosks that will be used for this this study and also, as I said, the replica of casino floor. Well, with that being said, this concludes my presentation, I'll be more than happy to answer to any questions you might have.

Peter Bowman:

Thank you so much, Dr. Benesch, we appreciate that fantastic presentation, you just showed us a lot of exciting stuff happening with Pepper, and we look forward to seeing the results. With that being said, we're gonna go ahead and open it up right now for some Q & A for Dr. Benesch. And in just a few minutes, we will go ahead and take it over to the robot lab team who can answer some more, the more technical questions concerning the Pepper robot, not relating to the research. So just looking at a few of the questions appearing here on the chat. Yeah. So we're looking through the Q & A right now, and I see a lot of questions about the Pepper robot itself. But not a lot of questions for Dr. Benesch, specifically. So with that being said, I think we're gonna hand it over to Cedric, to go through some of the questions. Cedric.

Cedric:

Yeah, so thank you, Peter. So really quickly, a few questions we had were related to the differences between now and taper. So for those who don't know, now is the older, small brother of Pepper now was released back into the eight. And it's still here, we actually just released the new version of now called now AI edition, which is very exciting. We launched it two weeks ago, so now we see here sit around, and still performing very well. So obviously, the main differences between now and Pepper is the size first of all; now is 23 inches tall, while Pepper is four feet tall. So that's one thing. Second thing, now is a legged human robot with three legs. Pepper doesn't have legs; Pepper is on her basis with three wheels that you might not see or; you can see it on the camera here; so that's how she looks. So very different from now. And the third and last thing about the design is obviously the touchscreen. Pepper does come with a touchscreen embedded on the chest, now doesn't have any screen; so it's mostly, you know, vocal commands and vocal interaction.

So those are really the main differences between now and Pepper in terms of design, and in terms of software, it's actually very similar, both now and Pepper comes from the same manufacturer. So Vancouver, TX; they are the same software ecosystem based on the operating system called melki. On top of that the robot is fully programmable, and you have access to, you know, the full SDK and full API access. Okay, if you want more information about now, feel free to reach out and we'll be able to give you more details about the latest edition of now AI edition. So that's about now and Pepper. How is Pepper being used in education; two main use cases. The first one is for, you know, a platform as a teaching tool for coding and robotics.

Once again, Pepper is a fully programmable platform, so it is used to introduce students to coding and robotics. Like for now, Pepper can also be used through for those who know choreographs software, which is a software that is included by default with the platform chronograph does look like, you know, scratch obliquely, for those who know, it's a drag and drop programming interface, where you get access to libraries of free program behaviors, your data sensors. And from a drag and drop type of thing you can create flow of interaction, dialogues, etc. And you can actually code in Python through chronograph.

Peter Bowman:

And just to expand on that a little bit, Cedric, the really fun part about Pepper. And going over to a little bit to the previous question, as well as the difference between now and Pepper, right, is the presence of Pepper; so not only are you allowing the students to learn coding and programming through Python, or the Android Studio platform, but you can make it a fun project as well. Right. So obviously, Pepper again, brings that presence to the table, due to its size. And so Pepper is also an excellent campus ambassador or program ambassador to any level of education, whether it be a high school STEM program, or a college research lab, et cetera. You know, you can have Pepper be that Ambassador to your program, and you know, travel, we have Pepper transport cases, for example, if you want to make Pepper a little bit more mobile. So it's just another layer to the interaction. And when I do want Dr. Benesch to be able to answer your questions as well. So I do see one question here on the chat that says, “Dr. Benesch, do you expect for people to be able to experience problems in understanding humans in noisy environments?” So Dr. Benesch? I'll let you take that one.

Dr. Benesch:

Sure. Yeah. Thank you. Um, well, this is a great question. I think there might be some problems. I mean, because, you know, I was like, Pepper recognizes voice, but I mean, if it there, you know, too many voices, it might be a bit difficult for prefer to maybe focus on one, we haven't had any issues with Pepper being, like that, you know, as specific you said, but I also have to say, like we are, we are now being very noisy. And it's like in a lab, so it's very different. But when there are people talking, I can say that Pepper will try to respond, even if you're not talking exactly to Pepper.

Peter Bowman:

Oh, the best part about that, too, is that again, it is more of a noisy environment, right,
Pepper does have that tactile option for you to be able to touch the tablet. And just a couple of things that you may or may not have been able to pick up on during Cedric's interaction, there's a few different indicators on whether or not Pepper is actually listening. Number one, it has lights on its shoulders, right. So if you notice during the presentation where Pepper was talking a lot about itself, right, those lights actually turned off, that indicates that Pepper is not listening at that moment. And then also at the top of the screen, it has text to speech or speech to text, right. So Pepper will actually show you what it heard. So you as the person interacting with the robot will know how to adjust the way that you're speaking so that the Pepper can actually understand. So those are two really important points. And Cedric, I'll let you continue on with some more questions.

Cedric:

Yeah, we do have tons of questions. Thank you guys. I'm not sure we'll be able to cover everything. Because you know, we actually pass the time. It's a 10; past 10. So yeah, I'm going to try to cover as much as possible, and obviously we'll be able to reach out back to you guys with the unanswered questions. I have one, “does the arms and hands have input sensing and good opposition for programming intelligent work task?” So the answer is yes, Greg. However, the main purpose of the bill is not actually to carry items, the arms and the fingers of the robot have not been designed for that; the arms and the hands and the fingers of the robot that you see have been designed to give a feel of a human like type of robots.

Again, focusing on human-robot interaction. If you are looking at robotic can manipulate objects, we do have other solutions here, such as a robotic-arms. We don't have any robotics on here for the demo for the webinar, but we can schedule with you another session to show you other options, where, you know, the human interaction is not really the focus, but maybe more than the object manipulation is one. We also have robots that can deliver items from point A to point B. So again, you are going to load items on the box and then send the robot to the other end of your hotel. Or to do in room service or in a bar or the cafeteria or restaurant; so this technology is available. We offer that but it's not going to be Pepper it's going to be a robot that is especially designed for delivery.

How can Pepper be used in teaching stem for high school? Great question. So we have actually many, many schools that are currently using Pepper, and as I mentioned before, Pepper is coming with choreographed software, which is a software that is extremely user friendly. For students that have no prior experience in coding or robotics is a great introduction to coding with drag and drop programming, as well as Python. If you want to go a little bit further for advanced users. We also have a curriculum designed for high school students that is also available for Pepper. And actually, we do have activities for any grade level, we also have a curriculum for elementary students to do storytelling with the robot, the robot is a great platform for that, as you saw, Pepper can talk out loud in multiple languages, Pepper can, you display information on the touchscreen, use body language, and because of that paper is a great, great storyteller.

Can paper handle multiple languages besides English? So again, the answer is yes. Pepper can speak up to 23 different languages, and I saw another question before, how would the robot know what to speak? So actually, it is not about Pepper, but the AI and Giants currently available, are not at this stage yet of recognizing a dialect or language. I am French, my wife is Colombian and I live in the US. So I have Alexa at home, it's in my kitchen, I come back home tonight, and I'm going to start speaking in French to Alexa, she is not going to understand anything, I need to manually switch her to French. So the way we do, we can do it with Pepper actually; we’ll trick it a little bit. And what we could have is a robot that can say hi, or hello in English, because everybody knows what Hi means. But then the robot can say if you want to speak another language, please select your language on the screen. And then on the touchscreen, you can have you know flags, and if I see the French flags, I'm going to click on the French flag, and then Pepper understands that she's supposed to switch to French language. So this is how we're going to be able to handle it. Okay. What do I have here? What's the payload? For the answerWe already talked about it. So it's not designed for that? What do we have?

Peter Bowman:

And so everybody, while Cedric is going through your questions, I do apologize, but we will not get to all the questions today, we will answer two more questions here. I just want to say thank you everybody for your time, I will be giving out our contact information here at the end. So anybody whose questions did not get answered during today's webinar, you will be able to reach out to us following the webinar and kind of follow up. Again, my name is Peter and Cedric will be answering two more questions before we conclude the webinar.

Cedric:

Yeah, two more questions. So some of you guys ask questions about, you know, what's available by default with the robots? So similar like for now, and I don't know, if you guys, you know, some of you might have been using now, the smaller robots, now used to have and has a mode called now's life, where it's basically a solitary mode, where now is going to come with some predefined equations. Same for Pepper, Pepper also has this type of mode. If you get the even if you get the Pepper with SDK without anything else, without the Quota Management System, you seem to have a purpose life type of solitary mode, where the robot is basically preloaded with I think, 20 plus question that you can ask to the robot, as well as few motions and a couple of dances. But other than that, you have basically the tools to program the robot to do anything else, basically, and that's number one.

And number two, another question. Which one should we pick? They are so many, and all of them are really great question, guys. So really, thank you for everything. And I have a question about from Clarence. So the question is, will we have to possess computer programmer to successfully operate Pepper? I own a four heart, that's another robot, a robot head, and it's limiting what we can do with for heart without having a computer programmer on staff. So great question Clarence. So again, as I was mentioning at the beginning of the of the presentation; again, you have two options, option number one, like four hearts and other programmable robots, you get the robot with the SDK, and then good luck. Okay, option number two; you want to have something that is more user friendly. In that case, you will need the content management system that basically allow you to use pre-skills already loaded on the robots. So basically in order for you to create or update your own content on the skills, which means create dialogues, or show another menu with your logo, and maybe some videos or pictures, this is like, as easy as using PowerPoints; really; it’s extremely easy and actually, we'll be able to train you on that.

And maybe the very last question. I know that also we have a hotel near Disneyland that is actually using Pepper. This question I believe is from Jonathan, we have been using Pepper at a hotel near Disneyland. She's very popular. But as we know, LG animated PowerPoint presentation, what are the capabilities of chat bots? So anything like a Google or Alexa device? This is a great question. And I believe if you're asking this question, it may be because you are not maybe aware of familiar with all the skills that are currently available with a content management system. But as an example, one of the skill of the content management system is what we call the conversational mode. That is using Google dialogue flow. So this is basically a chat bot that is actually accessible. And you can basically, you know, use it to, you know, create conversations and have the robots answer any type of questions. So yeah, so again, tons of questions. So thank you guys for your interest. We'll definitely get back to you with, you know, the answers to all of your questions.

Peter Bowman:

Yes.

Cedric :

Peter, you're taking.

Peter Bowman:

So thank you so much everybody. Again, I have been your host, my name is Peter Bowman, you can reach me via email. It's also in the chat, but my email is peter@robotlab.com. And again, my specialty lies in higher education. So anybody that's with a college or university, please feel free to reach out to me with any questions. To my left; one of the stars of this show is Centred , Vice President of Sales here robot lab. He is also our b2b specialist. So our folks at Disneyland that want help with their Peppers, Cedric will be able to help you out. His email is also in the chat and is cedricv@robotlab.com.

Cedric :

And by the way, if you're on a university or business, you can still email me, especially for district schools. And I'll be able to obviously, take care of you.

Peter Bowman:

And Dr. Nisi Benesch, who is our very special guests. Thank you so much for your time, and for bringing your research and, you know, speaking, speaking with us today, thank you.

Dr. Benesch:

Thank you, thank you for inviting me.

Peter Bowman:

And then off camera, we have Maria, who is going to come on camera now and say yep, she's gonna say bye to everybody. Thank you so much for your time. We appreciate it.

Cedric :

All right. Thank you guys. Have a lovely day.

Maria:

Thank you have a good one.

Cedric :

Thank you. Bye bye.

0 Comments
  • Sep 12, 2022 9:01:33 AM

Relevant Posts

Popular Posts