Forty years hardly makes a startup, but that’s exactly what Optimal Dynamics is. The New York-based company is seeking to flip the script on what artificial intelligence can do in a logistics environment, and it is starting with technology that has its roots at Princeton University, nearly 40 years ago.
Created by Warren B. Powell, the technology is unlike any other AI technology currently in use in trucking, says Daniel Powell.
Daniel Powell and his father, Warren, have taken the work Warren developed and tested over many years at Princeton, modified the technology and are now using it – what they call “high-dimensional AI” – as the basis for Optimal Dynamics’ entry into logistics.
Warren has now left Princeton to join his son as co-founders of Optimal Dynamics, with Warren as chief analytics officer and Daniel as CEO. The company on Thursday announced an $18.4 million Series A round of funding led by Bessemer Venture Partners, alongside returning investors Fusion Fund, The Westly Group, TenOneTen Ventures, Embark Ventures, FitzGate Ventures, and industry veterans John Larkin and John Hess.
High-dimensional AI is at the heart of the company’s CORE.ai product and is made possible by the use of approximate dynamic programming developed specifically for freight transportation. The company said this specific technology innovation will transition the logistics industry from real-time human decision-making to advanced automated decision-making at every level.
Daniel Powell spoke with FreightWaves on the trajectory of the company, the funding and the future.
Describing the company as offering a “decision layer” for logistics, Powell said that what is unique about Optimal Dynamics is that CORE.ai has been tested over many years in real-world conditions, including by companies such as Schneider National (NYSE: SNDR).
“All the code had been put into production, which is the really unique part of our business,” he said, noting that Norfolk Southern also utilized the original Princeton tech.
It wasn’t until the mid-2010s that technology advancements allowed the Powells to start exploring the next generation of code, using that original technology as the underlying support system.
“The technology is incredibly legacy-based,” he said. “We literally wrote the book on it.”
What Optimal Dynamics offers is a new “modeling framework.” Powell explained that CORE.ai – a software-as-a-solution (SaaS) offering – ensures users are able to automate decisions throughout the organization while receiving insight on how each of those decisions will impact other areas of the organizations.
“People forget that these operations are networks, they are not singular events,” he said. “It’s a highly complex operation and everything you do affects other [parts] of the operation.”
A basic example is a buyer that accepts a load that pays really well without understanding how that decision will impact other loads and could potentially put a contract at risk as assets get relocated. The same goes for the fleet manager that might reassign an asset without realizing that could leave a load – and customer – stranded somewhere else.
Powell offers a more complex problem faced by many fleets.
“Today, bids are mostly done in a vacuum state. They are scoped out and [bid] on as a separate entity. But they are not. If you take that bid … that is going to have an impact on drivers on other lanes,” he said. “You may need to hire more trucks or pull trucks from other accounts. Every time you pull pieces of the cobweb, others get pulled with it.”
That single bid can affect other lanes, on-time delivery percentage for additional accounts, hiring of drivers and even cash flow operations.
“We do all the hard decisions holistically, so we are always monitoring the entire company,” Powell said. “You can filter and view however you want … but [explained] in a very simplistic state, you can add new loads on one lane and see how other accounts benefits or other accounts suffered.”
While others may claim their technology can do similar things – and in some cases it can, albeit by pulling data from various systems – Powell said the reason Optimal Dynamics is gaining traction is because “the technology out there is just far too basic to drive the decisions we’re making.”
“We’re never analyzing a lane by itself. We’re always modeling the entire network in a holistic way,” he said. “It’s not just one bid, but 20 bids, and how do those bids affect the entire network.”
The CORE.ai product has been tested with companies involved in trucking, rail and energy, along with projects in health, e-commerce, finance and materials science, the company said.
“Bessemer is excited to partner with Optimal Dynamics to help drive their continued expansion. Logistics operators need to consider a staggering number of variables, making this an ideal application for a SaaS product that can help operators make more informed decisions by leveraging Optimal Dynamics’ industry-leading technology,” Mike Droesch, partner at BVP, said in a release. “We were really impressed with the combination of their deep technology and the commercial impact that Optimal Dynamics is already delivering to their customers. We look forward to helping them continue to build the future layer of supply chain intelligence.”