Manufacturers are looking to fill 3.8 million jobs between now and 2033, with the majority of the workforce retiring or missing key skills. At the same time, headlines are discussing how humanoid robots and AI are transforming everything. We are reaching a critical moment in time, where technology is the key to evolution, but it can be hard to determine what's ready to scale in the real world with publicity and demos.
That's what Ujjwall Kumar, CEO of Teradyne Robotics (the company behind Universal Robots and MiR), explored in his Automate 2025 keynote session. He's spent more than two decades in the industrial trenches — from General Motors to GE to Honeywell — and now stands at the top of one of the world's leading robotics companies. This makes him the ideal candidate to share what actually scales in industrial environments and why robotics is the future.
This article breaks down his powerful keynote session into actionable insights for anyone evaluating, deploying, or scaling industrial automation. We'll cover why you can't afford to wait, what's actually scalable, and what you should do next.
WHY INDUSTRIAL ROBOTS ARE NEEDED NOW
Let's start with the macro picture. According to a global trends report, the working-age population in developed economies is in steep decline, with a particularly sharp drop expected over the next decade. This means fewer hands to run factories. Fewer people are willing to do repetitive, dull, or physically demanding tasks. It also means a widening gap between the work and those who have the skills to do it.
.png "We have a labor shortage. Whatever workforce we have left out there, they will not do the kind of job we have made our workforce do in the past. – Ujjwal Kumar")
This is important because labor challenges aren't just about the quantity of workers and skills. It's also about our expectations for today and tomorrow's roles. The new and incoming generation isn't lining up for jobs that are dirty, dull, and dangerous. If we want to bring manufacturing back to the United States, we need to build factories and facilities aligned with shifting preferences.
We also need to support what’s anticipated over the next several years. For example, the acceleration of reshoring with new investments in semiconductors, energy infrastructure, and advanced manufacturing. Or, even the advancements in automation and technology like AI becoming more commonplace, human-centric design surging, and the mobile manipulation sector growing.
Future outlooks are what make automation solutions and understanding foundational today. We’re talking about real solutions that work. Systems that are safe, reliable, flexible, and accessible for those across industries.
But if we need it now, then why are we so focused on the hype?
HUMANOIDS VS. INDUSTRIAL REALITY: WHY FORM DOESN’T EQUAL FUNCTION
If you follow robotics news, you’ve probably seen videos and demos. Humanoid robots in homes, walking through warehouses, navigating disaster zones, and so on. It’s exciting, and the projections are as bold as the innovations. Goldman Sachs estimates the humanoid market could hit $38 billion by 2035. Morgan Stanley goes further, predicting over 77 million humanoid robots deployed globally by 2050.
But reality is much different than a demo in a controlled environment. Here's why:
- Battery life: Most humanoids in development offer 1 to 3 hours of operation. That's far below what's required for shift-level reliability in a manufacturing environment.
- Payload capacity: Today's humanoids typically top out around 20 kilograms. That's well under the requirements for many logistics or production tasks, especially in sectors like heavy machinery, packaging, or palletizing.
- Cost and complexity: Replicating human form introduces vulnerability, unnecessary joints, and reduced energy efficiency — all while adding minimal value for industrial workflows. Humanoids also require active control to maintain stability, making them inherently less reliable than wheeled platforms.
- Infrastructure readiness: As Kumar points out, most factories simply aren't set up for humanoids walking around. The basic infrastructure (floors, layouts, safety protocols, etc) isn't designed with bipedal robots in mind.
Now, this is not to say humanoids can’t find a place, eventually. Kumar sees compelling opportunities in elder care, hospitality, hazardous environments, and even defense.
But for manufacturing, "the form factor simply doesn't fit the process," Kumar says. "It introduces complexity and cost, reduces uptime, and fails to deliver the precision or specialization that task-specific robotics can already achieve today, far more efficiently and reliably."
So, the core takeaway here is that form must follow function, not the other way around.
.png "Don't try to build around the robot. Build around the problem. – Ujjwal Kumar")
WHAT'S ACTUALLY SCALING RIGHT NOW
Right now, the most common tools for manufacturing are autonomous mobile robots (AMRs), collaborative robots, and physical AI platforms. These three technologies are solving real problems.
What is an Autonomous Mobile Robot?
AMRs are mobile robots equipped with artificial intelligence to help navigate dynamic, unstructured environments in real time. They can avoid obstacles, move materials, and adapt to changing floor layouts and traffic patterns. And, most importantly, they do it all autonomously.
Kumar emphasizes how AMRs are beyond prototype. Companies like Mobile Industrial Robotics (MiR) have been perfecting this technology for over a decade. Today, AMRs are proven across industries, including electronics, logistics, food and beverage, and more. You can see many of these technologies live in the AMR Demo area on our show floor.
What is a Collaborative Robot?
Collaborative robots (also referred to as cobots) are designed to work safely alongside people. They can handle higher payloads, offer longer reach, provide greater stability, and follow a simpler path to deployment. Universal Robots pioneered this category and continues to lead with tens of thousands of collaborative robots deployed globally.
What is a Mobile Cobot?
Yes, a mobile cobot is what happens when you combine an AMR with a collaborative robot. It’s also how you get mobile manipulation, or the ability to create flexible, space-efficient automation setups. One example is MiR’s MC600 mobile manipulator, created in collaboration with Enabled Robotics and Universal Robots.
What is Physical AI?
Everyone has heard about generative AI. In industrial circles, the convo also revolves around physical AI, when artificial intelligence moves beyond analysis and prediction into physical execution. It is what helps humanoids learn, see, and adapt. But it’s also already powering robotics platforms today, from identifying components with high variability to learning and improving feedback to optimizing delivery routes and avoiding traffic jams.
THE 5 MUST-HAVES FOR INDUSTRIAL ROBOTS
Kumar discusses five non-negotiables for industrial automation that work in real-world manufacturing environments. If you’re evaluating robotic solutions, these should be on your checklist.
Safety and Reliability
This is table stakes. If a system isn’t safe to operate and reliable under real-world conditions, it won’t get deployed. Just like with equipment and machines, people need to trust a robot won’t cause damage or fail unpredictably mid-shift.
Real, Measurable Results
Manufacturers don’t (or shouldn’t) invest just to chase headlines or get buzz. It’s about improving quality, productivity, accuracy, and profitability. In a 2024 Universal Robots survey, those four factors were the top drivers for technology implementation. ROI needs to be quantifiable and quick.
Better Workplaces
This is one of the most overlooked points. The robotics and advanced automation technologies you see today are not designed to replace people. They’re built to equip them. They allow humans to take on more creative and rewarding tasks, eliminating the unsafe, repetitive, or unattractive ones they don’t want to take on. This benefit goes beyond day-to-day productivity gains, becoming a workforce retention strategy as much as an automation strategy.
Seamless Integration
While some operations do complete overhauls, the majority focus on modernizing their legacy systems or approaches. That’s why automation must be able to slot into existing workflows, tools, and infrastructure. If it doesn’t work with what’s already in place, it’s not going to scale. As you research solutions and talk to system integrators, be sure to ask about integration and implementation early on.
Fast, Agile Deployment
Manufacturing isn’t a static environment anymore. There are shifting product mixes, fluctuating volume demands, and evolving customer needs. Many large manufacturers are realizing they need to operate with the same agility as a small business to keep up. Instead of spending years perfecting an automation approach, it’s about testing, optimizing, and scaling. Your automation projects must be able to evolve as fast as your business.
Build Flexible, Scalable Automation with EcosyStems
One of Kumar’s strongest arguments was about how scalable automation gets built. It needs to be open and flexible, shifting to thinking about the ecosystem versus a single product or provider.
We see this on full display at Automate, where booths for one automation company will feature components or tech from another. These solutions are created by a mix of integrators, equipment manufacturers, software providers, and innovators solving specific industrial needs.
Kumar described the tech stack as three layers: standard, configurable hardware at the base; open APIs for both hardware and software in the middle; and an ecosystem of partners at the top. For Teradyne Robotics, that’s partnerships with companies like NVIDIA, Siemens, Honeywell, and many others.
This approach helps unlock innovation, deployment speeds, and integration. It’s the opposite of monolithic vendors or doing everything in-house. Instead, collaborating to enhance each layer's speciality and performance.
Real-World Example: Robot Palletizing
An example of an ecosystem coming to life is Universal Robots' PE20 Robotiq Palletizing Solution. This engineered system handles up to 20 kg payload at speeds up to 15 cycles per minute. It uses smart pattern generation and integrated vision guidance and fits seamlessly into existing end-of-line configurations with a compact footprint. It delivers fast ROI with intuitive programming that requires no robotics expertise.
Deployment and Strategy
Now that you know what’s hype vs. reality, it’s time for the next step. Here are some practical next steps depending on your role.
If you're in engineering or operations:
- Audit your current bottlenecks. Where are labor, quality, or throughput issues most acute?
- Evaluate potential solutions against the five must-haves framework
- Prioritize quick-win applications like material handling, palletizing, or machine tending
If you're in R&D or innovation:
- Map emerging technologies (physical AI, vision systems, mobile manipulation) to specific use cases
- Test with ecosystem partners, not monolithic vendors
- Build pilot programs with redeployment in mind from day one
If you're in procurement or strategy:
- Shift from "cheapest upfront cost" to "total lifecycle flexibility"
- Ask tough questions: How fast can we redeploy this? What's the integration timeline? What does your ecosystem look like?
- Demand interoperability and open standards
LET’S SOLVE REAL PROBLEMS TODAY
Whether you're leading a plant, managing a supply chain, or setting national policy, the future of manufacturing depends on scalable solutions that work today.
Ready to see these technologies in action? As the leading robotics and automation event in North America, Automate is the place to inspire, start, and scale your automation approach. Learn more and register for free today!
« View All Blogs
Privacy Settings