The Naked Truth about AI in Logistics

Highlights from a recent LinkedIn Live event co-hosted by Tony Wayda & Brian Carlson 

Much of what is currently promoted as AI is not new but enhancements of existing technologies, enabled by better data access and improved computational power. In this candid discussion, Tony Wayda and Brian Carlson reveal the Naked Truth about AI in Logistics.

In this LinkedIn Live event, Tony Wayda, Principal at JBF Consulting, and Brian Carlson, Principal at Cornerstone Edge, discussed the evolving role of AI in the supply chain and warehousing sector and explored how much of it is marketing versus factual AI applications.

Watch a recording of the 17 minute session here (stay for bloopers at the end)!

Key Takeaways:

• The field of Artificial Intelligence (AI) was formed in 1956, but this past year—almost 70 years later—has brought AI to the forefront for many businesses.
• AI's potential in warehousing is substantial, particularly in improving performance through labor predictability and transaction analysis; however, its full capabilities are still developing, indicating a gap between current marketing claims and actual functionality.
• Much of what is currently promoted as AI is not new but enhancements of existing technologies, enabled by better data access and improved computational power.
• When correctly implemented and trained, AI can provide significant operational improvements and insights. AI will change the way companies operate in the future.

Warehousing

In theory, AI can be implemented to improve overall performance, analysis, and software methods in warehousing. For example, theoretically, AI could assist in predictability around labor and look at the transactions that come in. However, while there is a lot of promise in this, the actual functionality is not yet fully developed.

We have seen AI being used to enhance forklifts driven by telematics in maintenance, system performance, and labor efficiencies. Recently, there have also been efforts made to increase AI capabilities in automation, conveyors, and other equipment, given that a significant amount of data goes through those systems.

As AI continues to improve in the coming years, the benefits and overall functionality of AI in warehousing will as well.

Telematics and Predictive ETA

Telematics has been around for a long time, and while there have been advancements, some of the capabilities currently being touted as AI are not new. Instead, there has been a great rebranding—everything that has mathematical formulations or analytics in it is now AI.

In reality, people are looking at pre-existing capabilities from a different perspective and utilizing the data more effectively.

Likewise, predictive ETA has existed at some level since the 80s and 90s, but it has recently become another buzzword. The biggest advancement in predictive ETAs is the large number of data elements available today, coupled with considerably better computing power.

There are now elements, especially in a cloud-based world, where you can gather information from within your existing systems and capture data from external systems or companies. This ability to collect data from so many different data points—IOT, real-time weather, real-time traffic, etc.—and process that on a timely basis is where AI will add real value.

The next advancement would be for the system to interpret the information and make intelligent decisions. For example, say an accident occurred on the interstate, blocking all lanes. The three routes affected would automatically be rerouted, the driver would be alerted, and a new route would display on the driver's OBC or handheld device.

Ultimately, when we stop talking about AI in broad terms and start talking about the capabilities of the systems, AI will lose its buzzword status and become something productive.

Software Examples

Despite a general overuse of the term AI, some companies are doing interesting work regarding natural language usage for digital assistance. Loadsmart has their system analyze data for them using natural language prompts. The system interprets the request and displays the result on the screen. A demo can be viewed here.

Oracle has digital assistants, the same concept as Loadsmart.

Greenscreen AI is delving into predictive pricing based on historical trends and contextual relevant data sets. They then pull all that data together to tell you what you should buy a shipment for, what you sell for if you're a broker, etc.

It is important to remember that when using AI technology, models have to be trained. Without training, models are unable to learn and grow.

Conclusion

AI has been around for a long time and will continue to be for the foreseeable future, but much of what is being touted as new AI is instead enhancements to pre-existing technology. These innovations have been made possible due to an increase in data points and greater computing power.

AI is incredible technology; as such, it is easy to get caught up in the hype. Separating the actual functionality of AI products from the hype and marketing will enable you to make the best decisions for your company.

AI holds great promise and will change the way companies operate today, but we are just getting started.