Ignacio Pedrosa, the applications lead for Novanta Drives, recently hosted a webinar about motion control in logistics robotics.
Ignacio Pedrosa, the applications lead for Novanta Drives, formerly Ingenia Motion Control, recently delivered a webinar about motion control in logistics robotics and gave insight into how to simplify the application development process. We followed up with Pedrosa after the webinar, which is still available to watch on demand, to ask his thoughts about the state of the robotics market.
The robotics market is growing rapidly, and the speed of that growth will only accelerate in the coming years. What are some interesting trends you’re seeing?
My answer to this question in 2023 is very different from what it would have been during the previous decade. We are now entering a time of huge change. Not too long ago, I would have spoken about the growth of mobile robotics industries during the past five years, especially in warehouse usage and last-mile delivery in the past couple of years. Or I would have mentioned how consumer and domestic robotics are slowly becoming a reality.
But in 2023, there is no bigger, absolutely game-changing core technology than AI. Throughout 2022, Dall-E, Stable Diffusion and ChatGPT were among the top viral topics on blogs and social media, but they are only the more tangible faces of a technological revolution that is just starting to emerge in the field of robotics. I think it is fair to say that when ’80s and ’90s movies amazed us with their technological visions of the future, we all thought robots would play a greater role in our societies by today. Robotics has grown quickly and steadily during the last decades, but not as fast as we had imagined.
But this is going to change with AI. AI enables human-like capabilities in robots, making possible everything from fully capable domestic assistants to versatile workers. It will affect transportation, medical care, commerce, education, security, the list goes on. The combination of robotics and AI has the potential to change our whole society, and the companies leading that change today are going to be the bigger companies of tomorrow.
What is a current example of how AI is changing robotics?
Industrial robot arms have been increasingly transitioning to collaborative robot arms over the past few years. The main reason is a robot arm alone might have the dexterity needed, but it lacks the intelligence and versatility than allows humans to perform sophisticated tasks. Hence, collaborative robots have found a wider area of application by supporting human labor to increase productivity rather than replacing it. But in November, Amazon launched Sparrow, an industrial robot arm featuring artificial vision and AI. Sparrow is capable of handling millions of different packages, even those that were never previously encountered by the robot. Interestingly, it does not use a collaborative robot arm as the main actuator, but rather an industrial one. That’s because Sparrow is not intended to help a human operator, but completely replace one. Is this the beginning of a return to industrial robot arms, thanks to AI? Time will tell.
With how fast robotics is changing, time-to-market is more critical than ever. From a development process perspective, what does this mean for end customers and manufacturers?
Developmental issues and errors cost time, money and lost opportunities. For end customers, it means a painful wait. It means not being able to simplify and improve lives, manufacturing processes, cities, etc. today because the technology is not ready.
Unfortunately, this is all too common. For example, a startup has a fresh approach to automating food delivery in a specific area but struggles to buy a fleet of autonomous robots to get the job done. This is because, although those robots already exist, goods delivery companies develop them for their own use, but they don’t sell the actual robot. This will probably change in the coming years, but today this startup company has to wait for a manufacturer to make it, or invest massive amounts of time and money to develop its own technology.
Manufacturers face similar challenges. If there is a market for a new technology, component or robot, offering a product ASAP is a must. Otherwise, the startup above will buy the robots offered by the manufacturer’s competitors. Even worse for the manufacturer, the competitor will probably shape the future trends of the specific industry by creating standards and patenting the technology. The first to colonize a market can impose its rules. That’s why development delays can be particularly devastating to manufacturers.
From a design engineer perspective, are there things you see that can help OEMs to develop their robots faster and more effectively?
My most important advice would be: Don’t start from scratch. Whenever you can start with components that already provide certain guarantees, use them. That’s true even if it looks like modification would entail more time than starting from scratch.
Unfortunately, we can be overly optimistic when forecasting potential problems and the investment required to solve them. The need to overcome these unexpected complications often comes at a time when the engineering team is already focused on another challenge, when the deadline has already passed, or with no budget assigned to something that decision-makers in the company thought was already finished. This non-allocated work almost always happens under pressure, and with the sinking feeling as developers that we have failed in predicting the issue. But the fact is, R&D is mainly exploring the unknown, and predictions are going to be wrong. Therefore, minimizing the chances of error at the beginning is always an investment that will pay off far more than you might expect.