Transportation/Logistics

Transportation/Logistics


Businesses have long struggled with how to move assets around the network in the most efficient way. There are so many variables to contend with: shifting demand, human error, traffic, fuel costs, weather, etc. With technological advances, the amount of data companies collect on a daily basis is astounding. The trick then becomes analyzing this data to gain meaningful insights from which business impacting decisions can be made.

Data scientists are able to apply advanced mathematics to answer many of these business questions which provide the insights to management necessary to appropriately maximize return on assets. Predictive logistic analytics is applied to predict machine behavior, human behavior, and even weather behavior. When companies begin to adopt optimization and predictive analytics into their logistical operation, they experience less mechanical downtime, more efficient routes, happier customers and higher stock prices.

Mosaic is uniquely qualified to help organizations glean insights from their data. We have more than a decade of experience designing and developing predictive analysis and decision support tools for NASA, the FAA, Boeing, Lockheed Martin, UPS and FedEx. Mosaic provides world class analytics consulting applying cutting edge machine learning and algorithm development techniques with unparalleled domain knowledge.


One of the largest North American Trucking Companies – Revenue Management and Demand Forecasting

  • Utilizing our flexible Rent a Data Scientist service model.
    • Staffing levels correspond with project demand and necessary skill sets.
      • Have access to a large team of data scientists.
  • Provide data science expertise and extra ‘brains’ to help solve problems / pursue projects as they come in.
    • Synthesizing market intelligence using third party, industry, government, and internal customer data.
    • Predict demand patterns for different market segments, regions and seasons.
    • Root cause analysis of seasonal demand patterns.
      • I.e. predicting demand patterns in the Southeastern United States for produce transportation.
  • The customer is able to better position their fleet to maximize revenue and profit with the data insights we provide back to the business.
  • Work collaboratively with their fresh-out-of-school data scientist and provide on-going training.

Link to full case study.


Largest Oil and Gas Manufacturer and Distributor – Machine Learning Inventory Imbalance Model 

  • Objective: Reduce instances of excess and insufficient product (gasoline and diesel) at over a hundred pipeline terminal locations.
    • Predictive machine learning model looking 15 days into the future.
  • The model allows distribution decision makers to have current situational awareness of inventory risks across the distribution network, improving productivity and facilitating quicker and better decision making.
  • Combines a time series model and classification trees to predict the risk of high/low imbalance for each product at each distribution terminal in the continental US. Daily predictions are generated with the most up-to-date data and presented to schedulers via a simple dashboard which integrates into their daily workflow.

Link to full case study.


Largest Global Express Shipping Companies – Custom Airport Decision Support Tools 

  • Improve operational efficiency with more accurate predictions on overnight package aviation operations.
  • Decision support tool running 24/7 at multiple airports providing situational awareness and demand predictions.
    • Gives decision makers awareness across all operational areas, including ramps, the ramp control center (tower), and the air traffic control coordination group.
    • Uses a variety of customer data feeds and complex prediction and optimization algorithms to compute and recommend ideal engine-starting and taxi times for the aircraft.

Link to full case study.