This article is sponsored by Arduino. In this Voices interview, Control Engineering spoke with Guneet Bedi, Senior Vice President and General Manager of the Americas at Arduino, to discuss the company’s approach to open ecosystems, rapid deployment, AI at the edge, and preparing the automation workforce for the future.
Editor’s note: This interview has been edited for length and clarity.
Control Engineering: Please tell us about your background and your role at Arduino.
Guneet Bedi: I’m the Senior Vice President and General Manager of the Americas for Arduino. I set up the B2B as well as the commercial operations for industrial and enterprise. Arduino is very interesting as a company because we’re the largest ecosystem of developers in electronics, with an estimated user base of over 33 million people globally. We’ve developed that over the last 20 years, and what we’re doing now is pivoting to create truly industrial and enterprise applications. Before joining Arduino, I spent about a couple of decades in large companies like Oracle and Cisco, and have been involved for the last 6 – 7 years in various Internet of Things startups, driving a similar kind of growth from our developer ecosystem into using the products in enterprise and industrial applications.
How do you define “accessibility” in the context of industrial automation today — and what are the biggest barriers that still remain?
Overall, there has been a lot of growth happening in industrial automation. A lot of tasks which were human, if you may, are now getting automated. One of the biggest concerns in this has been about vendors that don’t really create an open ecosystem. This forces innovators to pick a partner up front in order to design their automation solution, choosing an architecture up front as well. They end up in a place where they also have to hire talent that is very specific to that vendor.
When we say Arduino is opening up industrial automation to all, we think of it in two ways. The first one is to not lock you into any vendor — not even Arduino itself — by allowing a more standards-based approach. Secondly, we don’t believe in locking data. So, every time you adopt an Arduino solution in industrial, the data is yours. There are no license fees or restrictions for you to get the information you need for your operating and industrial environments.
And there’s actually a third element that comes into play from an accessibility standpoint. There are 40–50 million developers who have been trained on how to use Arduino in general, or how to use standards like IEC languages or C++. This gives entrepreneurs a huge talent base to create new automation applications, making innovation and growth accessible.
The industrial space is increasingly focused on digital transformation, but many organizations still struggle to move from pilot to production. What do you believe is missing from that equation — and how can open-source or modular platforms help?
We feel there are two main concerns for people to go from pilot or POC to actual production. The first one is to really make the business case work and make the ROI feasible. Open or open-source approaches allow you to make it very affordable and very cost-effective to do certain applications.
The second thing you need, along with affordability, is the right business ROI. And for that you need the automation deployment to be very rapid. That’s really what Arduino can help with. We’ve seen people save up to 30% of downtime, increase their efficiency by 10 to 20%, but do that at a fraction of the cost and at a very, very high speed. The combination of affordability as well as the speed of execution allows you to get to the business case, and once you have the business ROI, we see a significantly higher percentage of success going from pilots into production.
As edge computing, AI and cloud services converge on the factory floor, how can manufacturers avoid overengineering — and instead prioritize practical, scalable innovation?
I think with more accessible solutions and AI at the edge making applications really affordable, you don’t need to have expensive cloud platforms, which could also be complex to implement. You can truly do a lot, at the electronics level, with AI. A few examples are computer vision and vision-based production inspection, anomaly detection, condition monitoring, predictive maintenance — a lot of that can be done at the edge with a model trained in the cloud but implemented at the edge.
We’re seeing this makes applications real. My favorite example is with one of the largest furniture manufacturers in the country, Steelcase. They were literally using stopwatches and notepads to get insight into the efficiency of the factory floor. But now, with the combination of AI and open architecture, they can automate the whole process. Algorithms can tell them if a machine is working well or not, or if it needs maintenance. That really gives them the power to move from human tasks to automation.
Workforce readiness is a major concern across the automation ecosystem. In your view, how can industry leaders ensure that education, training and tools keep pace with the speed of innovation?
We see various statistics about millions and millions of jobs being displaced. In today’s world, this means that you have to reskill. That’s why Arduino works very closely with both industry and large university programs and community colleges. We have a significant portion of our business, as well as our installed base, in education.
We start as early as elementary school, but really the skilling and reskilling happens in high school, dual enrollment vocational programs, as well as in community colleges and universities. Kids need to learn a new approach which is not stuck to a vendor — like we said before — not stuck to one kind of logic. In industrial, for many years we’ve been sticking to IEC languages like structured text or ladder logic, but we also need to come to the modern world of Python or C, where a lot of automation tasks can be done much easier with new languages.
Beyond school, some of our students are 55–60-year-old automation process engineers who are learning how to do things in a significantly better and easier way. Reskilling is going to be an important part of the puzzle. Especially with the displacement of jobs, the industry has to come together with the education world and create compelling programs. Ours is what we call the Arduino Academy, designed to retrain and reskill industrial and process engineers on the new way of doing things.
Looking ahead, how do you see the role of open-source hardware evolving in industrial environments — and what new kinds of collaboration do you think will define the next decade?
One of the main aspects to keep an eye out for is how we can use a combination of human work and artificial intelligence to really automate factory floors and remote asset monitoring. Open source plays a very critical role in this, especially in industrial, because assets often last for many, many years. Unlike consumer, a 20–30 or even 50-year lifecycle is not that uncommon.
This makes it even more critical that you have an open approach, where openness allows multiple kinds of machinery and multiple kinds of technology ecosystems to talk to each other. Only then, in a more cost-effective and practical way, can the industry move to IoT or Industry 4.0, where these different ages and generations of machinery can communicate — which has to happen with open source.
I feel this will lead to a lot more consortia forming, as well as to multiple vendors collaborating with each other. We’ve seen examples of cloud vendors starting to work with hardware and semiconductor vendors. We’ve seen almost every vendor start to adopt OPC UA as a standard, versus being locked into their own silos. But all in all, I think open source will be critical, especially considering the extremely long lifecycle of industrial assets.
Arduino is the leading open-source hardware and software company in the world. To learn more, visit arduino.cc/pro
