Welcome everyone. Today we're going to talk about the rise of intelligence robotics, a vision for industrial and humanoid automation. And today I'm joined with Nakul Duggal. And Nakul, tell us a little bit about your role and your responsibilities at Qualcomm.
Good to see you Stephanie. I run the automotive, the industrial, and the robotics business. Been with the company for 30 years and really exciting time to be managing these industries, these markets with all of the opportunity ahead of us.
Well, this year has been a huge year for IoT and the automotive space, but 2026 is going to be about humanoid robotics, industrial robotics, AMR. So, let's jump into that topic. And I think that there's going to be a lot of exciting news that we'll dive into and we'll explore a little bit more about where Qualcomm is now and where we see the vision for the future. So, first let's talk about industrial and humanoid robotics. They'll be at front and center right at CES and to launch CES. We're going to see a lot of activity and a lot of announcements in that space. So, maybe give us a little bit of a realistic view on where we are today and where you see us kind of moving forward.
You know, we are super excited to be launching our own robotics initiative at CES. This is something that has not been going on for a long time, but we feel like we are very prepared simply because of how much has happened in the last decade in terms of AI and machine learning just becoming so mainstream and so democratic. You know, we started to invest in AI platforms and especially edge AI platforms many, many years ago. What we have seen as AI has grown is physical AI, which is essentially automation at the edge. First in the automotive environment, so driver assistance moving to automated driving and robo taxis. Then industrial automation, so more and more automation within our factories, and then eventually automation in the enterprise, in the medical space, in the service space. All of this requires you to be able to have a variety of technologies that have to coexist such that you can automate safely. You can automate locomotion, moving from place A to place B. Automate reasoning. So you want to be able to get the robot to be able to understand exactly what the task is. And then the most complex part, how do you automate in the physical environment? How do you move things? How do you manipulate? How are you dextrous with it? How are you careful with what it is that you're dealing with? All of that is converging now. And I think the next five years are going to be super interesting because the entire industry is trying to go solve this one problem.
So you talk about physical AI and I know the integration of artificial intelligence and robotics will be front and center at CES as well. You know, I think robotics is a priority for Qualcomm, right? So let's talk about that. Maybe dive into a little bit more about your unique role in this industry or this market space and where you see this playing out, you know, over the next year and maybe some things you want to dive into about some of the activities you're working on.
So, you know, what excites us most about this category is that it is truly an edge AI application. It's physical in nature. It has to work within constraints, whether constraints are power related, battery life related, weight related, safety related. But this is an application where you are going to the cloud to get your instruction and then acting. It's a real-time application, right? So that excites us because that's something that we have been working on for many, many years. We are good at that kind of stuff. The other piece that's super exciting is that it is by nature multi-sensor. So robots are instantiations of what some parts of the human body should be able to go to. So you should be able to sense, you have many senses, you have your eyes, you have your ears, you have touch, and robots have that same capability or that is exactly what is being designed. You also have the ability to reason. So if I want to manipulate my environment, I have to be able to understand the environment first and I have to be able to figure out the difference between A versus B. All of this has now been happening at scale in our industry over the last decade or so. So we find ourselves in this pretty unique place where the semiconductors that are needed to build these robots, we've been building that scale for quite some time. The infrastructure needed to be able to develop the AI models, the training for these models, the data collection for these models, that is all of the new capability that we are adding to our portfolio. But all of that comes together and we're going to go scale it starting 2026.
So you mentioned the human kind of characteristics that we see, you know, humanoid robotics are starting to display. Can you talk about how robots and humans are going to be collaborating over the next 5, 10 years? I know that's there's always a fear of that with workers and worker displacement. So maybe give us a little idea around how are we going to work together?
Yeah. So I think there are a couple of aspects to it. Let me break the problem into two parts. I think one is the question of robots and humans coexisting and working in the same environment. One thing that becomes really important is the concept of safety inside an environment where a robot and a human exist together. And it really just comes down to is the robot aware of the presence of a human? And if it is aware, what are the safety guardrails that you design the robot to be able to implement such that there is never any danger to the human. Today, if you think about robots in an industrial environment, they're in cages because they're moving at a certain speed. You don't want the human to be inadvertently in an environment where they could be harmed. And so for cobots or collaborative robots, you have to be able to have this very strongly implemented policy that ensures that the humans are safe. Now we design all of our chips for safety. In this case, we have to be able to design for understanding the presence of a human and what actions do you take in that space. So that is something that we are spending time on. To your other question about what happens with robots becoming more prevalent in our environments. Look, I think there are a couple of things going on. I think number one, the labor force is depleting. We don't have access to as much of labor as we used to in the past and a number of jobs that need to be done just have to be done in a more automated fashion and that is one place where as advanced AI starts to get deployed robots will play a much bigger role than they have. I think the other place that I see the robots being able to play a big role is in the service industry, whether it's in the hospitality space where you already see service robots or whether it's even in hospitals, in medical centers where you need to have certain tasks done, whether it's cleanup work, whether it's being able to tend on somebody, not to actually get involved but to be able to monitor a certain situation. A lot of different tasks that can be done by robots where a human is in the loop but from far away, from a distance. So tremendous amount of opportunity which still keeps the human involved but it improves the level of productivity and improves the level of care.
You mentioned about the need for work in terms of our workforce. I've also seen some stats out there around our just declining population and the need for automation and robotics as a result of that. So that's interesting to talk about. So given the need for edge intelligence, I know you talked about edge AI. So when implementing robotics, especially when it comes to industrial settings and even other use cases in manufacturing and logistics, what are some of the key technologies you think that are going to be really important for real-time processing? You talk about edge and we're really moving very quickly for low latency real-time data intelligence. Maybe give a little bit more details around.
So let me describe it to you exactly like the human body setup because I think it correlates really nicely. I think one is our senses, right. So when you think about a robot, the tasks are you plan the task rather, you sense the environment around you and then you act upon what your plan was. So what do you need to be able to go do that? You need your senses, you need your eyes, you need your ears, you need to be able to feel what it is that you are dealing with. And that for us translates to cameras, it translates to audio. It translates to be able to get feedback, tactile feedback from sensors that you have at the edge. And how do you process them together? So the concept of multimodality where you are taking multiple sensors, fusing them together and making a determination on exactly what the status is at a specific point in time where you are, for example, manipulating something, moving something, picking something up. So that is as real time as it gets. Is that what they mean when they say sensor fusion? So sensor fusion is basically taking multiple sensors and coming up with one specific understanding of what that specific situation might be. That is sensor fusion. Then there is the question of planning and reasoning. So for example, if I have to go pick the red ball and move it into the red box, the concept of understanding color, understanding a shape, understanding the difference between a box and a ball is something that's part of the reasoning, that's part of your frontal cortex. So that's your brain and so you apply a completely different AI model to it. A third aspect is motion. So if you have to move in your environment, you have to be able to have a sense of depth. You have to be able to know how far you are away from the object that it is that you're touching. So you have to simulate the environment in front of you and then be able to operate within that environment. How do you go do that? So the number of technologies that are coming together to be able to go pull this off. We've been working on these technologies for years because it's everything from location technology, camera technology, audio, the ability to fuse all of it together, do it in real time, do it within a power envelope, do it in a safe manner. That's what excites us about this because we frankly have been building for this for a long time and we will show at CES how quickly we were able to take platforms that we've been building for perhaps other applications and how easily they are able to transfer robotic capability on top of themselves.
