How AI and Machine Learning Are Changing Logistics and Freight Planning

For most of the past few decades, logistics planning has been a mix of experience, spreadsheets, and a fair amount of guesswork. Forecasting demand, routing freight, and managing capacity have all traditionally relied heavily on manual decisions layered on top of historical data. That approach worked (for the most part) until volatility became the norm.

The level of complexity logistics teams are expected to manage has skyrocketed in recent years for numerous reasons—such as a global pandemic, an e-commerce boom, and shifts in trade relationships, among many other factors—and the speed at which companies need to respond has evolved.

Now, artificial intelligence (AI) and machine learning (ML) are changing how companies move their freight. These technologies aren’t vague futuristic concepts discussed at trade conferences or innovative pilot programs happening only at leading Fortune 500 companies. Over the last several years, AI and ML have firmly embedded themselves into the supply chain, helping operators make faster, informed decisions in environments that change by the hour.

1. Continuous Optimization

Traditional freight planning models were always quite static and rigid as far as processes went. Routes were set in advance, and forecasts were built based on historical data. Any adjustments would happen reactively when something doesn’t go to plan.

Machine learning, and more recently, tools like generative and agentic AI, have changed that model. Instead of relying solely on historical averages, modern ML/AI systems continuously ingest real-time data about traffic patterns, weather conditions, fuel costs, carrier performance, and more. They identify patterns in that information and use them to predict what is most likely to happen next.

Although machine learning has been doing this for years, the difference today is the sheer volume of data being processed and utilized by ML and AI tools. With all this extra information, the result is a shift from planning freight movements at fixed intervals to continuously optimizing them based on what is happening in the real world in the last few days or hours. As a result:

• Shippers and carriers can adjust routes mid-transit.
• Bottlenecks can be avoided before they happen.
• Demand forecasts shift from rigid to dynamic.

These results translate into fewer surprises that cause supply chain hang-ups and faster recovery times when something fully unexpected does happen.

2. Smarter Forecasting

Forecasting has always been one of the hardest problems in logistics. Too much inventory ties up capital and space. Not enough leads to stockouts, missed deliveries, and unhappy customers.

AI-driven forecasting models ingest and process far more variables than legacy systems ever could. Examples might include seasonality, regional demand shifts, active marketing promotions, and external factors such as economic or social media trends. Using this amount of information in real time leads to more accurate demand planning that can be easily adjusted when something shifts.

More confident forecasts can create beneficial ripples across the organization, too. Procurement makes more informed purchases, warehouses plan labor more effectively, and logistics teams book capacity ahead of time (and at better rates).

3. Supporting Human Operators

Most real-world AI use cases are not about automating logistics workers out of a job. Instead, companies are using AI to inform decisions made by actual people by providing faster information about routes, carriers, or load configurations. At the end of the day, human planners still make those decisions. They can bring context that algorithms cannot (and may never) fully capture.

AI isn’t the lead actor in the play; it’s the stage crew. It makes its biggest impact by reducing manual work that slows teams down. Tasks like load matching and carrier assignments can be partially or fully automated, allowing planners to focus on higher-value activities and exception management.

“AI can’t and shouldn’t replace the people who understand how freight actually moves,” says Frank Crivello, founder and chairman of Phoenix Investors. “It’s merely giving them better visibility and faster ways to act on it. The companies that get and stay ahead will be the ones that use it to support their operators—not sideline them.”

4. Making Real-Time Visibility Actionable

Visibility alone doesn’t solve problems. Knowing where a shipment is doesn’t help much if you can’t act on that information quickly. With AI, logistics teams can see what’s going wrong and get suggestions on what they can do about it right now.

For example, if a shipment is delayed due to congestion, an AI-enabled system can automatically evaluate alternative routes, assess cost and service impacts, and recommend the best course of action. In some cases, it can even trigger those adjustments without human intervention, turning visibility into an active tool for managing freight in motion, sharing updates with customers, and improving service levels.

5. Building More Resilient Freight Networks

The past few years have made one thing clear: Efficiency doesn’t mean much without resilience. To this end, AI and machine learning are helping organizations model different scenarios and prepare for disruption before it happens.

Whether it’s evaluating alternate sourcing strategies, identifying high-risk lanes, or digitally stress-testing capacity plans, these tools provide a much clearer picture of potential vulnerabilities. When conditions change, instead of rebuilding plans from scratch, logistics teams can adjust existing models and move forward with greater confidence.

Where Logistics Goes Next

AI and machine learning are no longer experimental in logistics; they are the baseline. Of course, that doesn’t mean every organization needs to overhaul its systems overnight. In many cases, the most effective approach is incremental: starting with specific use cases like forecasting or route optimization, then expanding as capabilities mature.

What matters most is recognizing that dynamic, data-driven freight planning and forecasting are not only the future of the supply chain—they are happening now. The companies that adapt to this shift will find themselves better positioned to navigate whatever comes next.

About Phoenix Investors

Founded by Frank P. Crivello in 1994, Phoenix Investors and its affiliates (collectively “Phoenix”) are leaders in the acquisition, development, renovation, and repositioning of industrial facilities throughout the United States. Utilizing a disciplined investment approach and successful partnerships with institutional capital sources, corporations, and public stakeholders, Phoenix has developed a proven track record of generating superior risk-adjusted returns, while providing cost-efficient lease rates for its growing portfolio of national tenants. Its efforts inspire and drive the transformation and reinvigoration of the economic engines in the communities it serves. Phoenix continues to be defined by thoughtful relationships, sophisticated investment tools, cost-efficient solutions, and a reputation for success.