It's cringe-worthy that on April 28th, Newark Airport Control Tower had their computers using Windows 95 and floppy disks at one of the nation's busiest airports - it was indeed especially cringe-worthy for those landing that day. (Remember: with Windows 95 you could still enter <ctrl-alt-del> and wipe an entire disk drive.) During the terrifying 90-second air traffic controller blackout pilots were able to keep tabs on each other as they circled midair. Thankfully - even miraculously - no one was hurt.
Now flying out of Newark Airport is pure chaos. Meanwhile, the antiquated computer systems are getting swapped out for the newest and safest computer systems, and I'm quite certain that at least some small aspect of the safety system at Newark will have AI. Because everything safety seems to be trending with AI lately.
And as AI continues to trend and transform our world, it's essential to weigh out both its benefits and its challenges. Let's have a look at the pros and the cons of what is rapidly besetting us.
Pros of AI in a Paper Mill
- Reducing Workplace Hazards
AI fire systems like Bosch AVIOTEC have successfully used video detection in mills to rapidly identify smoke and fire. In addition to video systems, real-time monitoring systems track temperature, humidity, gas concentrations, or anything else you need and want, allowing for proactive intervention. Imagine if the Fires, Fatalities and Catastrophes in 2024 could be reduced by a significant percent!
- Predictive Maintenance for Safety and Efficiency
AI analyzes equipment data and predicts failures before they occur, thus preventing unnecessary downtime and equipment damage - by up to 15% in some cases. This unnecessary downtime, in turn, results in minimizing risks from injuries overall in the mill. That's the direction you want that snowball to roll!
- Improved Worker Safety with Wearables
Another option, AI -powered wearables, can monitor environmental conditions or even a person's vital signs, alerting to common potential risks such as chemical exposure or heat stress. Real-time AI alerts improve employee safety by addressing hazardous conditions promptly, which can work extremely quickly.
- Streamlined Safety Compliance
AI automates a lot of the (mindless) stats and data, reducing errors and audit prep time by up to 40%. Naturally, anything that's automated also reduces legal and therefore financial risks from non-compliance.
- Proactive Safety Culture
One thing AI is good at is predictive analytics, such as guessing how something might break or whether a machine might start a fire, based on past inputs of data. Predictive analysis also covers the potentially risky behaviors of people, such as hazardous forklift operating or signs of fatigue. However, if not carefully delineated, this can easily blur into some privacy issues, which we'll get to next...
The Cons of AI in a Paper Mill
- High Initial Costs
My oh my, can the implementation of AI systems cost a pretty penny. The upfront expense for hardware, software, and integration are all significant. Don't forget the ongoing maintenance costs. It's not pocket change.
- Data Privacy and Surveillance Concerns
Some people can wear AI-powered wearables and monitoring systems with no problems, and other feel like it's a too Big Brother-ish, monitoring their every movement, thus eroding trust. Each situation is different. Also when tracking such personal metrics, you need to ensure compliance with data protection regulations, which adds layers of complexity to AI tasks.
- Training and Adaptation Challenges
Clearly, mill employees require specific training to use AI efficiently and effectively, which costs time and money - and if your mill has an abundance of both (especially the money part), great! But if resistance pops up as well, this could quite easily hinder implementation.
- Risk of AI Bias and Errors
AI algorithms work great... until they don't. AI is based on its programming, yet it can produce biased or inaccurate predictions if based on flawed or incomplete data. This in turn, can lead to false hazards or missed risks. A good AI system will have this happen less and less, but you still need to be aware of these potential weaknesses. Furthermore, an overreliance on AI could increase these false hazards and missed risks in mills.
- Integration Complexity
Ask anyone who's worked in or with IT: integrating AI with existing systems can be a challenge and disrupt operations during implementation. Naturally, compatibility issues hopefully don't arise, but it ain't over till it's over. Your mileage may (and likely will) vary.
- Dependence on Data Quality and Infrastructure
A well-running AI system depends on quality data from sensors. Bad data in, bad data out, and when you have bad data or unreliable internet, this can limit performance. Also when you become dependent on AI for so much and you have the inevitable power outage or system failure, these can disable AI systems, leaving mills vulnerable to the very hazards they were designed to protect mills from, such as fires or chemical leak detections.
The extent to which each mill embraces AI is up to each facility. With this wide array of choices and options, as well as pros and cons, AI customization for safety should be relatively easy for each mill, ensuring it serves you without burden, and makes your world a better, safer place.
Next month we'll dive into Housekeeping, and the critical impact it has on pulp and paper mills. Stay tuned!
|
