Case Study

Predictive Empty Container Management for NileDutch

Project Background

Two major topics affected the container shipping industry in 2018: the trade tensions between the U.S. and China, the world’s two largest economies, and the Global Sulphur Cap regulation by IMO(International Maritime Organization). The new trade tariffs affected $34 billion worth of Chinese products imported into the U.S., with China taking countermeasures on U.S. imports. Slowing growth of global world trade volumes and an escalating trade war were expected to continue negatively influencing the container shipping demand in 2019, so cost-cutting has become critical for the container shipping industry. At the same time, carriers had to start preparing for the global 0.5% Sulphur cap, which would be introduced in 2020 and would affect up to 70,000 ships, according to IMO estimates. This regulation could lead to a 20-30% increase in total fuel costs, which would ultimately be passed on to customers.

Among the companies bracing for uncertainty in this market outlook is Nile Dutch Africa Line BV(NileDutch). Headquartered in Rotterdam, NileDutch is a major shipping company, specializing in container and break bulk shipping between West Africa and the rest of the world. The company has over 30 years of experience of working in Africa and has built a large network of offices and agents, supporting companies as they discover the opportunities the African continent offers.

In just a few decades, NileDutch has grown from a new market entrant to one of the world’s leading shipping companies focused on the Africa region, with a fleet of over 30 chartered container vessels and a capacity of 80,000 TEU. While being proud of its history, NileDutch has always been inclined to look forward, not backward. That is why the company is very ambitious in further growth through intelligent operational excellence by continually investing in its people, equipment, and technologies.

Problem

According to the Boston Consulting Group (BCG), repositioning empty containers in the shipping industry costs €13-17 billion per year – up to 8% of a shipping line’s operating costs. Further, there are additional costs for storage of empty containers and their maintenance, so the total cost of empty logistics is estimated by Transmetrics to be more than 12% of operating costs.

This is a huge industry-wide problem with many aspects to it, such as trade imbalances, long relocation time, high costs related to safety stock, as well as unreliable commercial forecasts, which are mainly based on the gut feeling of employees. In addition, there are two other important aspects that are applicable to many container shipping lines:

Using Excel for planning - Even nowadays, complex decisions for planning empty container logistics are typically handled manually with Excel and are based on the experience of logistics teams. Such planning is still possible since the majority of the container shipping lines own too many containers in reserve, which leads to significant cost inefficiencies.

No clear visibility on the costs - In many cases, the logistics team manages empty repositioning globally, while a procurement team controls vendor costs. With different types of costs in place, that can lead to situations in which container shipping lines know the total amount they paid to a certain vendor in a certain location, but the drivers behind these costs are not clear - for example, how many containers were moved, why were they moved, was it necessary to move them, if another vendor could do it, and so on. Thus, people making logistics decisions do not always have a full awareness of the cost impact of their proposals.

Even in these increasingly challenging market conditions, NileDutch was managing its operations very well with the help of good manual tools in place but it required significant time and effort from the Global Empty Repositioning team. As an ambitious company, NileDutch continuously seeks to improve its operational processes. That is why the client decided to empower its Global Empty Repositioning team by introducing a global end-to-end data-driven planning system. Among the main objectives of NileDutch were to reduce the logistics costs related to one of the biggest headaches facing liners, namely empty container handling and relocation, and to reduce the current fleet of containers that it operates, all while maintaining a high level of customer service.

Assignment

NileDutch sees data management, data analytics and making business decisions based on data as a critical enabler. That is the reason NileDutch has partnered with Transmetrics, the leading providerof predictive optimization and AI tools for logistics, in order to revolutionize empty containerhandling and relocation by utilizing AI, machine learning and intelligent algorithms.The goal for Transmetrics was to implement its predictive asset management software called AssetMetrics to further streamline empty container flows of NileDutch. Despite Transmetrics’ vastexperience in logistics, this project was the first implementation of AssetMetrics for the container shipping segment. NileDutch truly presented themselves as an innovative partner and a co-creatorof this new technology for ocean carriers by helping Transmetrics customize main AssetMetrics modules and co-develop them to fit the needs of the shipping sector.The two partners jointly defined the scope of the project and created a Kanban list of innovations and product enhancements. It was agreed that the software would provide daily rolling AI-drivenforecasts for the next 10-12 weeks based on the historical data and the external factors influencing the demand. The system also had to suggest the most optimal and actionable plan for emptyrepositioning, storage, repair & maintenance for the next 12 weeks, which takes into account all the related costs (including grading, stevedoring, gate costs, etc.).NileDutch involved several key people to make the project successful, including a project owner, project manager, one senior IT person for data and infrastructure, and one senior business user(Superuser). From the Transmetrics side, its main project manager was supported by a team of business analysts to address data transformations and a team of data scientists to developforecasting and optimization models. Apart from that, a steering committee consisting of senior executives from both sides was established to overview the general course of the project andevaluate outcomes.

Approach

The AssetMetrics implementation process included the following stages:

1

Setting up a secure data transfer

2

Cleansing and enriching the data

3

Building and testing the forecasting model

4

Building and testing the optimization model

Setting up a secure data transfer

Transmetrics offers an enterprise-grade VPN integration, which guarantees a high level of security for all daily automated data transfers. After Transmetrics completed the integration process, NileDutch could easily upload its data. As a result, the historical shipping data for the period of 2011- 2018 was available for cleansing and analysis, containing tens of millions of records, including booking routing and containers information, container scans, ship schedules, as well as ship and container reference data

Cleansing and enriching the data

A major focus of the data transformations for NileDutch was on determining which containers travel on which ship between every two ports. This information, however, was not straightforward to extract from the historical data, because there was some redundancy in different tables, and parts of the data in them was in conflict. Besides, a lot of the data was coming from EDIs from agents in different remote locations across the world, making it harder to maintain the data quality.

Nevertheless, the prior investment of NileDutch in digitalization, business process standardization, and a good TMS, were a critical factor for the success of data transformations and the overall project. To fully prepare data for analysis, Transmetrics, along with business and data experts from NileDutch, designed the appropriate business rules to prioritize information from the different tables.

After the business rules were in place, the team of business analysts and data scientists from Transmetrics cleaned and enriched the data by using algorithms to combine the full history of container scans with empty movements and bookings. This multi-step process ensured the completeness of the data, which was ready to be mapped to the ship schedules in order to determine which ports a container stops in between being loaded and discharged. During the data transformation process, Transmetrics also further improved the following data:

  • Booking routing – To cleanse the booking routing data, Transmetrics merged data from different tables, supplemented the voyage numbers and sequenced the number of feeders by data mining the information from the main voyages that immediately preceded or succeeded them.
  • Booking to scan mapping – To map booking routing to all container scan sequences, Transmetrics linked bookings to contracts, creating separate orders for each container (with the contract field linking them), treating a sequence of full container scans as part of the same order, mapping these scans to bookings, and then spreading this booking information on all scans in the sequence.
  • Ship schedules – Something typical for the shipping industry, ship schedules change occasionally due to weather, or other circumstances outside of a carrier’s control. To clearly distinguish planned versus actual dates and times of arrival and departure of ships for the optimization algorithms, Transmetrics prioritized information available from actual container scans about loading and discharge dates.
  • Booking information on repositioning – To enrich the data on bookings of empty container repositioning moves and mapping of container numbers to the bookings, with the help of advanced algorithms Transmetrics created orders for all empty movement sequences of scans
  • Information on voyage numbers and stops – The port sequence number information was crucial for linking scans, bookings and ship schedules. To transform and cleanse the data on voyage and stop sequence numbers, Transmetrics data-mined information on voyage numbers between container scans and booking routing information, carefully matching location and container data
Ultimately, Transmetrics developed a complete AI-driven data model, which automatically cleanses and enriches the data once the information is extracted from the TMS of NileDutch without affecting any existing processes for the client. The data from the model could be served for reporting purposes, for operational planning related to empty container logistics, and as inputs for forecasting and optimization algorithms.

Building and testing the forecasting model

Before implementing AssetMetrics software, the planning process of global repositioning of empty containers relied on the commercial forecasts from the agents around the world, which were mainly based on the bookings and personal business experience, and later were combined into one master forecast. This approach produced satisfactory results, but for the long-term it was not optimal because of its dependency on the manual inputs and human factor. For that reason, NileDutch wanted to introduce an accurate statistical forecast.

Having extensive experience in creating proprietary AI-driven demand models for logistics companies, the team of data scientists at Transmetrics developed a highly accurate forecasting model, which was trained on several years of already cleansed and enriched historical data. Furthermore, AI and machine-learning algorithms were used to continue improving accuracy with the new data coming in. As a result, the forecasting model was predicting the releases and returns of containers for NileDutch’s entire network for the next 12 weeks on a very granular level - per each day, port, ship, and type of container. It proved to be very precise even for the periods, such as the Chinese New Year, Christmas, and the like.

In addition to the automated statistical forecast, Transmetrics and NileDutch developed the functionality that allowed all the agents in various ports to collaborate and manually suggest their predictions (e.g., when acquiring a major customer, which cannot be data-mined from the past). Combining the powerful AI-driven statistical forecast with the years of experience of the NileDutch’s agents enabled Transmetrics to create the optimal forecasting model needed for the optimization module to plan the logistics of empty containers globally.

Building and testing the optimization model

Before partnering with Transmetrics, NileDutch already had in place good processes for planning empty container repositioning, but they were time-consuming since many tasks had to be executed manually with the tools such as Excel. Complex decisions for loading and unloading empty containers to ships, repositioning them to other ports, storing them in other depots or repairing containers were taken based on the extensive experience of the Global Empty Repositioning team of NileDutch.

Such a process was manageable because the company had many containers in reserve, but despite NileDutch having skillful professionals, it is impossible even for the best planning team to calculate the optimal, lowest-cost plan. In order to do that, one would need to consider a huge amount of inputs such as how many containers would be needed in the next 12 weeks in every single port of the network on few different continents, the schedule of each ship, the free capacity of each ship and how it would change over the period, the cost to transport the containers between hundreds of possible routes, the cost to store containers in each depot, how many containers you store at the moment, which of them are within the free tier and for how many you already pay for storage, the gate costs and stevedoring costs in different ports, how many containers your customers will return in each depot over the period and so on. If some of the inputs change, the planning needs to be recalculated. Such calculations are beyond the capacity of a human brain but they could be performed with AI technologies, and that is what AssetMetrics software does.

The global optimization module was co-developed in close collaboration with NileDutch to meet all the requirements of the container shipping companies. The Transmetrics’ team started from analyzing the necessary business rules, inputs and parameters. A major focus was put on an entire set of logistics costs, which must be considered to produce a meaningful optimization.

While working on the costs, Transmetrics’ teams of data scientists and business analysts identified several data topics, which had to be addressed first in order to provide the best inputs for the optimization model:

  • Repositioning cost - Necessary data inputs were cleansed (combined sublocations, price unit differences, duplicate rows, additional rows with charges to the main cost, mixed text and value cost cells, etc.) and when required, the data was mined from similar cases based on the closest port and by setting constant values (currency, depot, sublocations, values, container types, supplier codes, etc.). All the adjustments were implemented only after a deep discussion and confirmation from the client.
  • Depot storage cost - To cleanse and enrich the data for depot storage costs (e.g. information whether the storage and costs are calculated per tier or per TEU, some zero values, etc.), a trace back in tariff reports was made. When a trace back check was not possible, a constant value was set that the client agreed with.
  • Gate In/Out cost - For the cases when only gate-in or gate-out cost was available for the location, Transmetrics created a detailed report with the suggestions for costs based on machine learning and AI, which were validated by the client and later used in the optimization module.
  • Stevedoring cost - Before proceeding with cost calculations, Transmetrics had to compute the formulas in the Tariff column and to unify supplier names according to the supplier code. In general, stevedoring costs were based on complex business rules, so Transmetrics developed a custom automated data transformation process that considers all such rules.
  • Other costs - To enrich the inputs for repair costs, grading costs, and reference data (depots, ships, services, voyages, schedules), the Transmetrics data science team created reports with suggestions on how the data should be completed, which were accepted by the client.
    Since all of the client’s costs are dynamic, Transmetrics developed an automated data transformation process. It extracts, cleanses, and imports all types of relevant cost data related to the container logistics from the NileDutch’s financial system into the optimization model. In addition, both partners developed the screens where users have the ability to manually update each cost.

After addressing all the data challenges related to the logistics costs, Transmetrics continued building and testing a powerful global optimization module. This module takes into consideration the results from granular rolling forecasts and every single related cost as well as a status of containers, ship schedules, and other inputs. All the new optimization features on top of the existing module were developed based on the specifications from NileDutch. As a result, AssetMetrics optimization suggests the most optimal and cost-effective global plan for managing empty containers in a given situation. It is presented in several user-friendly screens, functionality and layout of which was also designed together with the NileDutch’s experts. With such user interface, the client’s planning team can view what is suggested in the system for every port, ship, or service every day. The client can also use the provided reports to understand how the suggested plan was executed in real operations

Client: NileDutch

A new automated planning system from Transmetrics allows NileDutch to make more thorough and data-driven decisions about the management of empty containers on all levels (station/depot, regional, global) and to have better visibility on the data and business processes.

AssetMetrics software is set up in a way that every day, in just a few clicks an optimized empty containers plan is re-calculated for the entire company, 10-12 weeks ahead of time. To achieve that, a new incremental historical data set from the client’s transport management system is extracted, transformed and added to the Transmetrics model. It is then used to build highly accurate rolling demand forecasts for the next 10-12 weeks. The forecast results, along with the corrected costs, are used in the optimization module to build the global plan for empty repositioning, storage, repair and maintenance for the whole network for the next 12 weeks at the lowest possible cost.
Now, to make the most optimal decisions, NileDutch can plan its empty equipment flows with AssetMetrics based on all cost inputs and other relevant business variables.

After the implementation project, NileDutch tested the system and benchmarked software suggestions to the results from dispatchers. The client confirmed that the benefits from AssetMetrics have been better than initially expected, the product is user-friendly, and that its
entire team was comfortable using it daily as a guidance in decision making, so the project proved to be a success.

AssetMetrics has demonstrated significant operational improvements for NileDutch. Namely, the total costs for managing empty containers were reduced. In particular, storage costs have been impacted very positively alongside the improvements for other logistics costs. The AssetMetrics system has also helped NileDutch to reduce its container fleet size.

There were several additional benefits from using AssetMetrics software. It allowed NileDutch to get rid of manual planning in Excel (Excel still remains one of the most widely used supply chain management tools). AssetMetrics also enabled the company to have “one version of truth” since now all the relevant data from different logistics and financial systems is shown in one place. NileDutch employees from all over the world could begin collaborating through the system, making them feel more connected than ever before.

AssetMetrics provided a lot of efficiency that saved a significant amount of time in the client’s daily business operations, allowing NileDutch employees to focus on further improvements. For example, NileDutch is using data aggregation reports from AssetMetrics on a daily basis, in particular about daily stock levels with prices, which helps to control the storage costs significantly and go into
details if spending is high. Besides, it helps to monitor storage facilities and calculate what is the most optimal free tier for each facility.

Transmetrics and NileDutch also identified a number of next steps to further increase the benefits of AssetMetrics solution. For example, the functionality about cabotage management, as well as local maintenance and repair strategy is in the development. Transmetrics is also planning to release a convenient management reporting dashboard, which would include cost reporting and an easy graphical interface showing how the supply chain looks like for the next 3 months (forecast for certain locations compared to supply and actual bookings).

Testimonials

We had been looking for a partner who could help us streamline our empty container flows for a long time. There are different IT solutions on the market, but we chose Transmetrics over other providers because they proved their technical expertise and also demonstrated a deep knowledge of the logistics industry and container shipping in particular, which is very difficult to find.

I can say that we made the right choice and NileDutch is a satisfied customer. Transmetrics is a very professional and reliable partner that has throughout the project delivered very good results. They have great project management skills, and also a very dedicated customer approach, which have been perceived very positively by NileDutch. After implementation, Transmetrics is still giving us very good customer service, being open for discussions on product enhancements and being always available for questions and corrections on the data or on the logic where it is needed

Altogether, we are very happy with the nomination of Transmetrics as our partner in empty equipment flow management and we can recommend their services to any other logistics provider in the network.

Carlo Zaalberg

Global Director Logistics

Case Studies

Case study

Predictive Asset Management in Trailer Leasing Business

Read full case study
Case study

Predictive Linehaul Planning in Express/Parcel Business

Read full case study