A.I can look at data, pull information from outside sources and reach conclusions that are beyond the capabilities of any human fleet manager.
Photo: Pexels – Google DeepMind
As is so often the case these days, you’re going along trying your best to get through the workday, and suddenly there’s a whole new buzzword everywhere. And you’re supposed to stop what you’re doing and devote time and attention to understanding. And then you’re supposed to start planning for how you’re going to inject whatever the buzzword is into your day-to-day business and become more efficient and profitable and so on.
And who has time for all that, right?
The latest buzzword in trucking now is “A.I.,” which, of course, is an abbreviation for “artificial intelligence.” And if you’re like most Americans, all you know about A.I. you learned from The Terminator:
At some point in the near future, the computers become self-aware, realized we’ve screwed everything up, and decide to kill us all. And to be fair, that’s not really that far-fetched an idea. When A.I. suddenly broke into the mainstream earlier this year, there were some pretty wild stories along those lines. But let’s put aside the pending annihilation of mankind by killer robots for now and look at what this new computing technology can do for you in your job, and for trucking in general in the coming years.
Now, for the record, I really don’t know much about AI myself. But, I was lucky enough to talk to someone who does. That’s Craig Scott, an expert in digital transformation and the founder of MFGx, a manufacturing software and software solutions integration company that uses Fuuz, its cloud platform, to help truck fleets and other customers manage telematics in data in maintenance, logistics, inventory, warehousing, and distribution.
During our conversation about Fuuz, I went off topic on AI and asked Scott to help me understand why it’s suddenly being touted as such an important technology for the next generation of fleet management systems.
Machine Learning Versus Artificial Intelligence
For context, Scott began his explanation with the concept of “machine learning,” which was itself an industry buzzword a few years ago.
Machine learning, Scott explained, has been around since the beginning of the computing age. In fact, he noted, it is one of the most basic tasks computers do: Take different sets of data, compare them, and draw some very basic conclusions about what that data is telling you. A prime example in trucking, he said, would be your maintenance management systems gathering and comparing data that shows that fan belt life on your trucks is about 60,000 miles. As more and more data comes in confirming that figure, the system will recommend changing those fan belts before they reach 60,000 miles.
That’s certainly useful information. But, bear in mind, a machine learning computing system hasn’t really “learned” anything or reached any conclusions on its own. It simply looks at raw data, crunches the numbers, and draws some fairly basic yet invaluable conclusions about what that data is telling you.
AI, on the other hand, does have the capability to look at information beyond mere, in-house, data and concludes as to what plan of action should be taken. That’s because AI can pull information from the internet relating to its tasks and use that information to improve those tasks in ways very much like a human would do.
Craig Scott, founder of MFGx, says that Artificial Intellegence goes far beyond computers merely crunching data and presenting recommended actions.
Photo:P Fuuz
As an example, let’s say you have a predictive maintenance program with empowered with A.I. One day while searching the web, the AI engine picks up on news stories announcing that the Commercial Vehicle Safety Alliance’s annual Safety Blitz is coming up in a couple of months. And that this year, CVSA will be targeting lighting systems on the commercial vehicles inspected during that week. Taking this information, an AI maintenance system can begin immediately placing an emphasis on lighting system maintenance across your entire fleet – without any human input at all. Likewise, the AI system could pick up information that CVSA will be focusing on driver safety (as it did this year) and could automatically begin sending out safety videos on driver communication platforms with warnings that this type of enforcement is coming soon.
Let’s go back to those fan belt failures for another example. We all know that fan belts fleet-wide simply failing at 60,000 miles isn’t a realistic scenario. Things are way more nuanced than that. One brand of fan belts might not fail until they reach 70,000 miles running in one part of the country. But that particular brand of fan belt is also likely to fail at, say, 53,000 miles if it runs X number of miles in another part of the county – in mountainous terrain, for example.
Now, a machine learning maintenance system, over time, would probably reach those same conclusions. The difference is that an AI maintenance system would determine those outcomes much sooner. But it would also then make the necessary ordering, inventory, and maintenance recommendations to address the issue without any human decision-making required. Moreover, the AI system could then adjust how trucks are assigned to specific routes to maximize the life of the components on the vehicle. And not just fan belts: Every component on the truck.
And this is just on the maintenance side of things. Imagine how powerful AI could be in terms of efficient freight and logistics management. And its benefits in autonomous vehicles are obvious, too. In fact, it’s quite possible AI could be the final piece of the technology puzzle that brings autonomous trucks into everyday use soon.
This is a level of detail, understanding, and action items that is beyond even the most gifted human fleet manager. And, when you look at those kinds of determinations and operational decisions being made by a computer, it can be a little scary to consider. This is why I think it’s safe to assume there will be a mandatory human manager review, confirmation, and approval process in place with these systems. AI will tell you what it’s learned and what conclusions it has drawn from that information. And then recommend a course of action based on that data, with more data detailing why those actions will be beneficial to your operations.
It’s heady stuff. But it’s also easy to see why AI will be such a powerful tool for fleets in the future. Instead of taking raw data and trying to reach conclusions, the AI system will do it for you. And on a much deeper level than any human ever could. And probably in ways we don’t yet understand.
Credit: Source link
