During a technical discussion at the 2023 SME conference Eric Winsborrow, executive vice president of corporate strategy at Wenco, explained that the data is only as valuable as the results it can garner. “In order to have a skill set around something realistic on the asset health side, you need a number of skill sets, you have to have machine learning, you need subject matter expertise; otherwise, what’s the point of the AI?” he said. He went on to say that finding someone that has experience with machinery knowledge and an understanding of the processes and data analytics is rare.
“Good luck finding someone who does all three,” he said. Another problem Winsborrow identified regarding AI was data overload. With information coming in from everywhere, deciphering what is actionable and not is especially vital if you are trying to implement predictive maintenance. “Now we have sensors on everything and what the hell do we do with that?” he said. On the flipside, the corporate strategist noted that sometimes the sensors don’t always work.
“And they certainly aren’t predictive in their nature. They actually tell you that the horses have left after the barn door is closed, so there is OEM sensor issues as well,” he said. “All of that leads to issues with trying to get to root cause.” Changing the way a company interacts with its AI can change results from reactive to proactive and aid in root cause analysis “AI, done right actually helps resolve a lot of these issues or at least puts a dent in all of these different issues,” he said. For Wenco facilitating the proper use of AI meant bringing on Montreal-based Rithmik Solutions and their advanced analytics. “What I liked about Rithmik was they came from tech resources, asset health, all the founders are from Matricom and knew how to connect devices, their initial use case was mobile mining equipment AI,” he said.
Using data collected by the miner, Rithmik will generate rapid analytics, visual trends and dashboards. From there, Wenco can collect real-time data utilising Hitachi’s numerous deployments globally. Winsborrow then provided some real-world case studies of the enhanced AI analytics and the results it attained. Using a slideshow, he explained that an African mining company had 370 over-stroke events annually. “The issue with that not only is lost productivity; obviously, a machine has to stop, but some person from the maintenance department has to go down to the pit, that is as a safety issue,” he said. Using predictive AI, Wenco was able to identify several issues. The first was the slack adjuster not braking where it was intended to brake.
“Another issue they found was that the RPM sensors failing, and actually getting some false positives around slippage where there was no slippage,” he said. “So the traction control system kept on trying to apply brakes to one side, and the poor operator was fighting, fully burdened or not, was fighting the braking system.”
The AI also found that the diesel in the trucks was not being used at the right temperature, resulting in US$70,000 of wasted fuel annually. Aside from the cost, the incorrect temperature burn was also an environmental issue. “The problem with not burning diesel at the right temperature is you don’t just get regular GHG emissions, you get particulates and other types of toxins that are very bad,” he said. So you’re creating a real sustainability issue because you’re not burning it at the proper temperature.” Wenco is also building a cloud platform that will provide IoT connectivity across Hitachi’s new truck fleet. “It’s being deployed out in Australia this year; we’ll bring it to the Americas late 2023, early 2024,” he said.