Watch Now


The tools keep getting better: McLeod’s applied data science initiative

Photo: Shutterstock.com

Last year, McLeod announced a new data science initiative which seeks to help our customers turn the information they have into more actionable insights and push that to the point of decision-making for users in many roles throughout their company. On Monday afternoon at the McLeod UC2018, speakers Randy Seals and Josh Jones reported great progress with the initiative.

One of the new ideas is the concept of a composite index they use to calculate for their customers called the Rate Assessment Window, or RAW score. This new composite metric seeks to quickly answer the questions “should we take this load?” and “at what price?”

There are obvious questions to be considered in order to answer that question, including: Is it profitable?
Is it good for our network? Can we cover it easily?
ake the contract price or counter offer? Hw much do we need to get? 
ht does the market tell us today?

“This is the most exciting thing I’ve heard about from a trucking perspective as far as I’m concerned,” said Seals. It’s all about applications for LoadMaster and AI, the Machine Learning Process.

They’ve been gathering the data, cleaning it up, and discovering applications from the data.

One of those current applications is using the analytics of load selection in real-time: providing direction choice by showing what is likely to happen, what should happen and the result. Using existing data supplemented by external data set. “Time is our only asset,” said Seals.

You also have to look at costing results, lane definition and defining the network. “Defining your network is where you want to be. This is where you save your money. We’re going to tell you what your operational costs are. It’s all about probability forecasting and all in real time. At the time of the tender, at the time of order entry, EDI acceptance, etc.,” said Seals.

“There’s two types of loads we’re talking about. There’s forward yield, where it’s going and network yield, where it came from.”

Josh Jones, CEO of StrategyWise, has partnered with McLeod. Jones is also managing partner of StrategyWise, a 20-employee consulting firm in Birmingham that helps corporations and foundations implement data science and big-data projects. Their projects all surround the goal of building capacity to understand current customer behaviors and operational patterns, or being able to predict those likely to occur.

“As a data science team we help companies figure out how to bring AI and big data into your fingertips,” said Jones.

“Using the load evaluation tool uses the idea of forward yield. It uses predicted revenue, cost and time with proposed revenue which is predicted cost and time and compare that to the next several moves of predicted cost and time as a way to track total and time to aggregate your total forward yield,” he said.

“Even for small decisions you make, the computer is calculating millions of small permutations every time. If you do this thousands and thousands of times, the tool gets better and better.”