(What can I say? I’ve never been much of a sports guy.)
The roof is 282ft high at its center and covers eight acres. It consists of three moving panels and one fixed piece. Two of the panels slide under the fixed panel, then the final panel rotates around the stadium and into place until the four panels resemble a stack of plates. This means the stadium must be capable of quickly, reliably and precisely moving 11,000 tons of steel in winds up to 40 mph blowing from varying directions.
Recently, the roof received a major retrofit, courtesy of a partnership between JMP Engineering, Rockwell Automation, Gerrie Electric and Cisco.
At nearly three decades old, it was due, as Dave McCormick, engineering manager for the Rogers Centre explained. "I always equate it to this,” he said, “Not many people are still driving a car they had 20 years ago."
ENGINEERING.com had the opportunity to discuss this unique project with Andrew Lyng, manager of JMP’s Toronto branch and project manager for the retrofit, and Steve Szamocki, JMP’s executive vice president of sales and marketing.
How does this project compare with others you’ve worked on?
Andrew: This project was unique; the Rogers Centre roof was the first movable roof in the world and after speaking with a lot of mechanical engineers, we learned that no two movable roofs are the same. Compared to other jobs, this project stands by itself: it’s not automotive, it’s not food and beverage and it’s not general industrial, which are the markets we generally work in. It was challenging as an application—moving a roof—but once you break the project down into smaller portions, there’s a lot of similar technologies and techniques that can be brought to bear from those other industries.
Steve: There was also a bit of an IoT component to this project, in terms of the industrial network. Everything is a hundred percent Ethernet now, whereas 29 years ago, when the roof was put in, that wasn’t the case. That said, I think the most significant difference was the size and scale of the entire project.
Did the project require the development of a unique OT network and control system or was the system adapted from another application?
S: Our goal was to avoid providing a customized, one-off solution that’s not supportable. So, we worked with Rockwell Automation and Cisco Systems on providing the backbone using Stratix switches and components from both companies to build a system that was safe, secure and reliable. The motors, drives and redundant PLCs now all operate in a very orchestrated fashion over Ethernet.
A: We used standard technology but applied it in a unique fashion to meet the requirements of this particular project.
You identified the engineering challenges of this project as the schedule and logistics involved, the size and weight of the roof and designing for wind shear. How did you address those challenges?
A: The secret ingredient in all of this is being a member of the CSIA [Control System Integrators Association]. When we tackle a project, it’s essential that we over-communicate with our customers and our partners. The number one reason that projects fail is a lack of communication. The whole project started for us in January 2015 and ended June 2016. It was a real exercise in planning and project execution.
S: Compared to an automotive, food & beverage or petrochemical upgrade—anything from the process automation world—you can’t build your risk mitigation plans in the same way. When you have a public venue with a rigid schedule of events, your planning has to flow around big rocks in the calendar. That means you don’t have the usual consecutive blocks to manage your time and the blocks are much shorter because any time there’s an event going on, you might have to remove all your equipment from public view. There was a lot more planning and logistics involved.
I understand the retrofit reduced the number of required operators. How was that accomplished?
S: The old system was installed in the ‘80s, so it used a lot of very basic technologies—switches, no feedback into the control systems, that sort of thing. But because you’re moving 11,000 tons of roof, you need to be very careful. Previously, they had approximately six people scattered throughout the roof’s structure to monitor and verify that the roof was doing exactly what it should be. With the new system, we’re able to access smart diagnostics with cameras and sensor confirmations for roof positioning and lots of built-in redundancies. They still have a small team of people up there to ensure safety, but the roof can be opened and closed with a sole operator.
The new system also increased the roof’s speed by 46 percent. How was that increase achieved?
A: That was calculated based on where the roof ended up on the old system, after 25 years of use. The control system was the last piece of the puzzle for refurbishment. We restored the system back to its original design specifications while also making it a bit faster. Before we did the retrofit, the opening and closing cycles of the roof were approximately 45 minutes each. Now, they’re able to open the roof in approximately 24 minutes and close it in 21.
S: The IoT network also lends itself to more precise communications between the drives and motors, which we upgraded as well. It’s really several things that account for that performance improvement.
A: There’s also the skew. There are two roof panels that straddle the entire stadium. The design specification stated that each side of those panels couldn’t be out of alignment by more than 50mm [2 in]. If they get out of synch, then they start putting more pressure on the rails and start slowing down. We were able to use Ethernet-based encoder technology—both absolute and laser—to maintain the skew typically under 10mm [0.4in] per side, at full speed. Once you have that high degree of control, which, as Steve said, came from the upgraded communications, we were able to maximize the performance of the roof.
S: If you think about the two sides of those panels moving in parallel, skewing less than 50mm is a big deal. Especially when you consider things like wind shear, where one side may be affected more than the other. We had to have dynamic braking that could be applied to one side or the other and constant communications between the encoders to keep things running smoothly in parallel. Otherwise, if one side gets more skewed than the other, you’ll end up with some big mechanical issues.
Are the fault tolerance, self-diagnostics and reporting capabilities of the new system applicable to manufacturing or other industrial operations?
S: The underlying technologies we used are directly applicable in many industries. The roof is being controlled by a completely redundant PLC system. With the advent of Microsoft Surface tablet technologies, we have people using those as mobile platforms for monitoring and control. That’s something we’re seeing more and more in manufacturing: people not being tethered to their desks.
A: If you look at it from the operators’ point of view, they have redundant controllers and redundant Ethernet going into a Stratus fault tolerance server system. From the server out, we also used technology that notifies personnel about alarms and other status changes. The roof automatically communicates to the people who need to know: if it’s moving, it tells them; if there’s a problem, it tells them. Cisco provided the security and the reporting was taken care of through Rockwell’s FactoryTalk View SE software.
What was the most rewarding part of this project?
A: For me, what brought it home was that over the couple of months we were commissioning that roof, we opened and closed it 60 times. That was during baseball games and other events, so there were 50,000 people in the stands. The local media is always on the lookout for the roof. “Is it going to open? Why isn’t it opening?” We were able to fully commission that roof and nobody even knew we were there. For a project of that magnitude, that’s a real achievement.