Exclusives : The Climb to the Top of the Smart Factory Data Pyramid

The Climb to the Top of the Smart Factory Data Pyramid

The smart factories of the future are set to be considerably more efficient. The destination on the road map from 5G to 6G isn’t necessarily a shock in that regard. How to get to that point might be, though.

The True Backbone of Smart Factories

For example, the interconnectivity of sensors and robots (and human operators) has led to a productivity increase of 2% at the United Kingdom’s recently launched first 5G smart factory. The robotics themselves won’t exactly be responsible for the greater projected gains moving forward, though.

Rather, it’s the data resulting from the automation that will end up furthering the projected efficiency, according to Jonathan Hou, CTO at Pleora Technologies, a sensor-networking solutions provider in markets like industrial automation. The greatest long-term positive impact as 5G rolls out will come from how collected data is leveraged, he said.

“It really starts with how to automate something, then adding efficiencies to get to a system that uses machine learning to analyse data and make performance improvements,” Hou clarified.

“For example, in an inspection system, if there are ongoing errors or defects, machine learning can use data and analytics to make automated adjustments to correct issues. This adds ‘self-improvements’ on top of an automated system. Accessing this data, be it over 5G or 6G, is really the backbone for these technologies.”  

Next-Gen Technologies Making a Difference Today

After mechanization, electrification, and automation comes the fourth industrial revolution, whose defining characteristics include greater interconnectivity between devices. As we move through this revolution, Hou believes a variety of technologies will contribute significantly.

Technologies and their use cases include:

  • Augmented Reality (AR), for example in the form of glasses, which overlay digital information on what the person wearing them is seeing.
  • Asset monitoring to check on the condition of materials and products.
  • Digital twins to simulate outcomes over the course of a given asset’s life cycle, from how it will connect to other devices to how long until it will need to be replaced.

These examples all leverage data in some meaningful way. While the mind may tend to connect Industry 4.0 more with robots on the production line, even those robots take advantage of data in how they’re automated.

Consider an automated Heating, Ventilation, and Air Conditioning (HVAC) system, for example. HVAC systems have been thrust to the forefront during the Covid-19 pandemic because of the need to optimise airflows and filter the air. Automated HVAC systems are installed in factories today, but they also figure heavily in the ecosystem of projected smart factories of the future.

While the initial slowdown hit the HVAC industry hard, there’s reportedly been a quick recovery based on high demand. Even before the pandemic, the global Condition Monitoring Equipment sector had been valued at $1.70 billion USD in 2017, with a growth rate of 5.56% between 2018 and 2023.

Ultimately, that’s what smart HVAC systems do: Monitor conditions, pushing for greater efficiency, including from a maintenance perspective. In the case of BrainBox AI, the first fully autonomous HVAC system, artificial intelligence (AI) gets leveraged for additional benefits.

“Using AI, we’re able to improve many aspects of the efficiency of buildings, including hitting sustainability goals, carbon footprints, improving overall occupant comfort, which is a really big deal,” said Andy McMahon, BrainBox AI’s Channel Sales Director, who reported BrainBox AI can improve the percentage of time a building stays at the temperature the tenant set it to by 60%, leading to total energy savings of 25%.

“It’s very low or no-touch… We’ve got [a team], that watches the building data 24/7. And as things change, we alert the building owner in real time of anything that we’re seeing. That’s mostly to check and balance what we’re doing as far as the AI goes… but, maintenance-wise, from the software perspective, that’s all on us.”

Keys to Success of Industry 4.0 and Beyond

It’s a pyramid of sorts according to Matthew Sallee, who runs the commercial and multi-family business units at Motili, an HVAC solutions provider with a nationwide presence in the United States. Data makes up the foundation. Collected data needs to be analysed. Once it’s analysed, you can take action. As you take action, you can optimise it. Then you can automate, which is the summit.

“There’s a lot of data and you’re looking at data that’s coming across at microseconds typically. Just floods of data coming off the components… and I think one of the challenges is not going to be so much handling that flood, because both 5G and the future of 6G will have adequate pipelines to handle that stream of information,” he said.

“[Because of the projected higher bandwidth] the network’s going to be able to handle the flood of data, but it’s going to be such a giant amount of data that it’s really going to be how do you process the data and what do you do from it. That’s going to be the trick.”

The greater projected bandwidth alongside ultra-reliable low latency are two key 5G wireless network benefits which smart factories are projected to employ. The subsequent capture of huge amounts of real-time data theoretically leads to improved mission-critical decision-making. Consequently, machine-learning can be applied. That’s just one step according to Hou, though.

“Machine learning is a manual process,” said Hou. “You’re training an AI model and deploying that on a machine. The next step is self-learning systems with ongoing, continuous improvement. In a factory setting, that requires equipment and systems efficiently ‘talking’ to each other and sharing data. To me, a smart factory is machines communicating to machines in an automated manner, and providing decision-support capabilities for humans.”

In McMahon’s view, complications arise from the integration of different hardware controllers, which are used to “run” buildings. Each has its own manufacturer and protocols.

“Over the years, as far as integrating those, essentially making those controllers all talk to each other under one platform, that’s been kind of the holy grail,” he said.

McMahon suggested the endgame may include the use of AI in those controllers to get them to work together, likening buildings to organic beings all the while.

“You can’t just take one aspect and try and control it and then say, ‘Okay, it’s efficient now,’” he added. “You have to look at the whole things as a body, and each part is important to take care of.”

Feature image courtesy of Lenny Kuhne (via Unsplash).




To reserve your ticket please fill out the registration form