AR Data Sciences
This Month's Q&A: November 2020
Iván Rodríguez-Conde, Ph.D.,
Department of Computer Science
U.A. Little Rock
“ALEXA, TELL ME about the Internet of Things.” That’s probably an interview for the future, but for now we prefer to hear about IoT from an expert in the field. As a researcher, Iván Rodríguez-Conde specializes in the connection between humans and machines, a relationship that has been heating up for decades. And with more devices than ever available to stake claims on our hearts and minds, we sat down with Dr. Rodríguez-Conde to find out just where this relationship is heading.
I have to confess that when I first heard the term “Internet of Things,” my first thought was, What a dumb name.
Yeah, I know, but there are a lot of things going on behind that term. And it's not only about appliances or, let's say, consumer devices. It’s a technological reality that has already begun to impact our daily lives, but it’s going to transform everything from medicine and healthcare to more traditional sectors, such as agriculture. And it’s going to have an extraordinary impact on manufacturing.
Until recently, the Internet has been associated with a mere network of interconnected computers that allowed us to communicate and access multimedia content. But that classic concept of the Internet has completely lost its validity. Since the appearance and rise of mobile devices and the subsequent appearance of smart devices, the Internet has grown exponentially, incorporating millions of new machines and electronic goods capable of capturing, processing, and exchanging data.
We’re definitely witnessing a fascinating time of rapid technological development in which innovations such as mobile devices, 5G networks, or machine learning are converging and subsequently pushing the expansion and evolution of IoT even further.
Last year, I was able to attend the Consumer Electronics Show, and I can tell you those innovations are already here to change our daily routines for good. I’m talking about things like Internet-connected appliances that will warn us when we run out of milk or TVs that we can use not only to watch our favorite online content but also for our workout, being able to estimate how many calories we burn with that workout.
Let me drop back a moment and ask what attracted you to this field.
I’m from Galicia, in Spain, and I went to school at the University of Vigo there. And when I finished my bachelor's degree in computer science, I started working as an intern in the SI6 Computer Graphics and Multimedia Lab. The projects that we had were all basically aimed at the development of graphics-based multimedia interactive tools. That's how I started getting interested in the relationship between machines—in this case computers—and users.
That led me to a project—this was in 2009—of creating some kind of interactive application that could allow the user to design a cabinet in an easy way. In Spain, it is very typical to create custom cabinets for your home. But it's kind of tricky because typically the designs are done using CAD software, which is very technically demanding. So my main goal was to come up with something that my mom or I could use to design the cabinet for our home, and then just send an order to the cabinet manufacturer, shortening the whole manufacturing chain.
What I came up with was a mobile application that provided the user with a touch-based experience—the user could, just with a finger, interactively design the cabinet. That was my first contact with mobile IoT work.
Did you do that as part of your Ph.D. work?
No, but let’s say that was my first research project in the lab, and that led to my Ph.D. One of the greatest problems of the users when they start a new cabinet design is having an idea. It's very complicated for many people to face what I call the “blank-sheet issue.”
So what I wanted was to make kind of a “recommender system” that could suggest several proposals to the user—several designs based on their particular needs. Let's say 20 percent of a user’s clothes are jackets, 50 percent are trousers, and 25 percent are T-shirts. That person has a style that is completely different from, say, a businessman. So the distribution inside his cabinet is going to be pretty different from the distribution inside a businessman’s cabinet. A big goal of my Ph.D. thesis was to come up with an algorithm that could automatically generate the distribution of the cabinet, allowing the user to customize the cabinet for his own wardrobe.
Once I successfully defended my Ph.D. dissertation, I started working in a company in the private sector as head of development of the mobile department. What we did was develop tons of different and tailored mobile applications.
For various businesses, or just for various uses?
For various businesses. We made, for example, a social network for travelers just to share different routes. We made mobile applications for logistics. We also made another mobile application for labor risk prevention. So yeah, all mobile applications but targeted to different domains.
After that, I went to work in a research lab that did work for the automotive industry. That’s where I started getting involved with what we call embedded devices, which are typically low-powered special-purpose computing systems. For example, I was involved in an autonomous driving project in which we developed amazing software such as a mobile application to control the steering wheel of the car just for testing purposes, or a lightweight 3D engine for digital instruments clusters.
I worked for that lab for almost two years, and after that I joined a group of people who were creating a small startup. They wanted to create something very disruptive, at least from my point of view—a mobile-based technology that allowed the user to use his or her mobile application from any smart device connected to the same network. The cool thing about this was that the app was capable of adapting the graphical user interface and their interaction to the target device.
Which meant, for example, that you could have your Facebook application on your mobile device with the mobile graphical user interface. But if you launched your application on the TV, you would see the same application with the proper layout. Or say you want to open your Gmail but you need to use a keyboard and a mouse to feel more productive—in that case, you just plug your mobile device to a screen and use your mobile device as a computer. No matter the target device you’re interacting with, it's running on your mobile device. The idea was to put everything you need in something you already carry every day in your pocket.
Nowadays, there are different alternatives in the market that are trying to do the same thing, but they’re mostly focused on one specific operating system or one specific operating platform. What we were creating back then was something generic for any platform, from any operating system. It was pretty cool—and then I got a job offer from UA Little Rock.
I was going to ask you how you made the leap from Spain to Little Rock.
Well, it was strange. I was working in that startup company, and one day I received an email from my college mentor. Fortunately for me, my mentor had finished his dissertation in Valencia, and Dr. Carolina Cruz-Neira, then head of the computer science department at UA Little Rock, knew a lot of people from University of Valencia. She was looking for someone for her department, and she sent an email around letting people know she had a position open.
And the people from Valencia sent that email to my mentor, who forwarded it to me. I did a Google search on Carolina Cruz-Neira, and said, "Sure. I want to work for her." I arrived in Little Rock in September of 2019.
Let’s talk about the history of IoT. Before we were dealing with all of these consumer products, what was going on in that field?
It was pretty much in the scientific domain, most of it related to robotics. And robotics is still the force that is leading the push for IoT.
In my research for this interview, I read that the concept of smart devices was discussed as early as 1982 with a modified Coca-Cola vending machine at Carnegie Mellon University. It could tell you when it was about out. It could tell you whether the drinks were cold enough. It was an “Internet-connected appliance” nearly 40 years ago.
Well, I can tell you many of the innovations that we have right now were all defined 20, 30 years ago. The main problem back then was that they didn’t have machines powerful enough to implement those ideas. Their workstations didn’t have fast processors, enough RAM, nor powerful GPUs. But they had great ideas.
In terms of consumer devices, probably the first devices that came in good in that big box labeled IoT is basically smartphones. A lot of people assume that the iPhone was the first smartphone, which is not right. There were earlier mobile phones that were able to connect to the Internet and allow users to browse the web. But the iPhone, introduced in 2007, was the game changer.
So are there certain milestones in the development of IoT?
I wouldn’t say milestones as in, “We first have to get to this milestone and then go for the second one.” More than milestones, I would say there are challenges right now for the short-term and the midterm. And there are a lot of people working in parallel to try to solve those challenges.
The first and most important challenge, I think, is security and privacy. Right now, everybody is using Cloud-based platforms, because they provide huge power, especially for machine learning. But the problem with these Cloud platforms is that in order to have these phenomenal intelligent systems, you have to fit them with a lot of data. And most of that data comes from the user, like bank accounts and other personal data, and that’s one of the biggest concerns from the user side.
The second challenge is the interconnection, the connection itself. Right now, we’re talking a lot about 5G, but we’re not there. The 5G that the carriers are selling us is not the 5G we were promised a couple of years ago. It's not the same speed.
At least from what I’ve read about 5G, the current speed is 2.7 faster than 4G. And what we were told from years ago is that 5G would allow almost real-time interaction between two devices connected through the Internet, which would mean, for example, that you could control a robot remotely from your smartphone. That would be something incredible. I mean, think about what we could do in the health domain with that. But unfortunately, we are not there, and I think it's going to be a major challenge.
Another thing that we’re probably going to need to address at some point is that we have this Internet with only PCs and smartphones and smart TVs connected. But in a few years, we are going to have a huge Internet of Things, which means we’re going to have tons of different devices. Right now we’re using a pretty old protocol, TCP/IP, and that protocol was rated for just a few devices.
So probably someone has to come up with a new protocol that can connect all of these coming devices. Because my guess is that we’re going to run out of IP addresses.
What do you foresee showing up in the next few years in terms of IoT?
Something we’ve just started to see in the smart devices, especially in smartphones and tablets, is machine learning. For example, in our Gmail we get suggestions for how to finish a sentence. That’s machine learning, and in the short term I think it’s going to be pervasive in our everyday lives.
I use Alexa every day to check the schedule for my day, or to check the weather. Well, in the future, what we are going to use is not just a device with Alexa. Alexa is going to be with us all day, because Alexa, or Google Assistant, or whatever the name, is going to be with us in multiple devices. It's going to be in our earbuds. It's going to be in our cars.
So anytime during the day, you could just say, "Alexa, tell me this"?
That’s right. And that’s going to change the paradigm of how we interact with machines. You will interact with machines just as you talk with your wife. And that, from my point of view, is going to be the major change, because the more natural the interaction with the machine, the more people—the larger number of people—will be able to interact with them. That means my mother or my grandma are going to be able to use a computer, to use a smartphone, with no restrictions. To some extent, the machine is going to be transparent for the user. They'll just be interacting with another entity, which just happens to be virtual.
When will this happen? How many years?
That’s a good question, and I don't have the answer because right now there is a lot of fragmentation in the technology domain. Google has its ecosystem, Apple has its ecosystem, Amazon has its ecosystem. There are some cases where they try to give these ecosystems some kind of compatibility, but that's tricky for the user because they have to have the skills to configure their devices to make them compatible. That's the main problem that I see right now.
So give me some examples of things we’ll be able to do easier when all of this takes place.
Well, my opinion is that we are going to be a little bit dumber in the future. Because this virtual intelligence is going to act as our mother, right? Just think about what we do right now. We cannot even remember the phone number of a friend or relative anymore. All we know are contacts. And if I have someone that reminds me that today I have a couple of meetings, why should I care about remembering those meetings?
The other thing is that we are going to be connected to others more virtually than physically. Nowadays we use social networks, and we can see photographs from other people. We can share videos. But in some years’ time, especially when we have 5G, by using VR and AR we’ll be able to talk with a virtual presentation of another person right in front of us. We’ll probably be able to have a coffee, a virtual coffee, with that person as if he or she were physically at the same table.
I think that's going to replace a lot of physical interactions, because it's very hard for a lot of people to find time to go out for a coffee and hang out with friends. Everybody works a lot of hours during the week, and sometimes they have to force themselves to actually go out and have that coffee with a friend. The coming technology will allow them to be in contact with the other person, and they won’t need to spend as much time as when they hang out physically.
Sounds like this is going to make us almost slaves to the machine.
That's what a lot of science fiction authors say. And based on what I'm seeing right now about how people use social networks, I don’t think that’s a crazy thing; I think it's physical. I don't want to be pessimistic, but I can tell you that the more I know what's behind those kinds of technologies, the more I personally try to be disconnected. For example, I don't have social networks. I don't like that kind of interaction. I prefer physical interaction and seeing someone face to face.
But I understand that social networking is something that will be around for a long time. And probably it's going to be worse for the next generations. There are a lot of kids who were raised using those technologies and having those kinds of interactions. So what can I say?