Any questions? Feel completely free to contact us

Transportation & Logistics Industry

By the Numbers: Transportation and Logistics Data

The global logistics market is one of the largest industries in the world, comprised of 10% of global GDP and with a value surpassing $4 trillion dollars. The transportation sector is the largest and fastest growing segment, with an annual growth rate of 7% annually since 2011 — this segment alone is expected to generate $3.8 trillion in revenue in 2016. The large umbrella of “global logistics” is comprised of a wide variety of transportation:

  • Air freight: Global value of $70 billion;
  • Sea freight: Global value of $54 billion;
  • Road freight: Global value of $2 trillion; and,
  • Passenger transport: On the rise with over 800 million passenger cars in use globally.
As expected, all of this movement of people and cargo on a global scale results in an enormous amount of data. If any industry is in a position to take advantage of Big Data analysis, it’s the transportation industry.

Benefits of Transport Data Analysis: Cost Savings

The predictive analysis of logistics data can be used to transform the way companies do business, particularly in terms of cost efficiency. For example, in the road freight and transportation sectors, fleets of vehicles often reach the tens of thousands — all of which require constant maintenance. Instead of defaulting to preventive maintenance, transport data analytics enables companies to perform predictive maintenance. Vehicle data gathered from sensors feeds raw data into a predictive analytics solution, which then cleanses, formats, and models the data to determine which components are the most likely to underperform or stop working before they exhibit symptoms. This type of analysis allows technicians to repair/replace components early instead of performing expensive post-damage repairs.

Benefits of Transport Data Analysis: Operational Efficiency

By linking historical activity data with consumer profiles, economic indicators, and geolocalized market data, logistics and transportation providers are able to predict demand with increasing accuracy. This allows them to anticipate daily volumes, optimize delivery routes, and allocate resources accordingly to deliver the service more efficiently, ultimately increasing customer satisfaction.

In addition, transport data analysis excels at discovering bottlenecks, particularly those derived from inefficiencies and a lack of optimization. The logistics industry is largely driven by economics: fuel cost, security measures, time to delivery, air vs land, supply chain reliability, domestic distribution networks, offshoring, and so on. There are a wide variety of factors that can impact profitability, all of which need to be taken into consideration in order to achieve optimized efficiency. The disparity & complexity of these factors means that predictive analytics software is, realistically, the only option capable of harnessing Big Data to produce real-world business solutions.

Benefits of Transport Data Analysis: Dynamic Pricing

The plethora of factors that impact logistical efficiency also impact end-product pricing. A change in one computational ingredient (e.g., increased fuel cost, security-related shipment delay) can have a profound impact on the overall shipping cost and, consequently, the product price. Price determination should be malleable and based on real-time cost data. Predictive analytics solutions are capable of factoring in cost-sensitive components and using that data, often combined with external dimensional data (e.g., weather patterns and transport time), to accurately predict an optimized price. This method of dynamic pricing ensures that your logistics operation balances competitive pricing with actual shipping costs.

Using Dataiku DSS to Perform Transport Data Analytics

Dataiku Data Science Studio (DSS) is a powerful predictive analytics solution that allows users to connect to a wide variety of data, quickly clean that data (in the GUI or with code), and enables them to creatively factor-in datasets to create predictive models. These models can be configured and tweaked, as needed, in order to visualize your own unique business scenarios. The analytics process uses a rich machine learning interface to empower your company to build predictive services based on past and incoming data.

Transport Data Analytics: Collection and Visualization

It all starts with data collection. To maximize value creation, logistics providers need to access and combine unevenly structured data sources, such as: historical activity, CRM data (e.g., orders, contacts), machine-to-machine data (e.g., truck sensors, RFID chips), geolocalized open data (e.g., trends, points of interest, geopoints), and maintenance operations records.

Dataiku DSS uses powerful ETL and data wrangling features to facilitate the data cleansing process. Processors (e.g., geocoding, IP addresses) and connectors (e.g., Openstreetmap) help to enrich geographical data for visual insights. Whether you are building an automated delivery optimization system or trying to understand bottlenecks in your logistics network, DSS will allow you to use and compare a broad range of algorithms and build the best solution for your needs. If standard algorithms don't do the trick, DSS also lets you integrate custom scripts, libraries, or models in Python or R.

Featured Use Case: Optimizing Chronopost's Logistics processes

Chronopost worked with Dataiku DSS to improve on-time deliveries during peak activity times. By geo-aggregating historical delivery and parcel retrieval data, Chronopost was able to develop an application that calculats an ease-of-delivery score for each address. This score-based approach enables them to flag risky deliveries and meet their delivery deadlines. Due to the increased operational efficiency of their network, Chronopost is able to decrease package delivery costs and develop new commercial offers.