Steinbeis experts co-develop a forecasting tool
The Steinbeis Innovation Center for Research in Transport and Logistics is working alongside TLT Berlin as part of a 12-month research initiative sponsored by the State of Brandenburg. The aim of the project is to develop a model for predicting variations in parcel and shipment volumes in the general cargo market.
With general cargo, consignments can weigh anywhere between 30 kilos and 2.5 tons. The main task cargo companies face is how to group lots of small consignments with larger deliveries and still ship them over major distances. The market for general cargo has expanded rapidly over the last 15 years. One of the most noticeable effects of this change is that customers are looking for service providers who can “cover all the bases.” This is because customers are eager to work with a single freight forwarder or parcel company, somebody who can look after all their cargo requirements. Another important need relates to short delivery deadlines, which are usually within 24 hours. To successfully address both parts of the equation – full service and quick delivery turnarounds – suppliers need sophisticated infrastructure and comprehensive networks. As the clients in these networks are based in all kinds of locations and their logistical requirements entail distributing shipments to recipients in similarly diverse locations, the providers of their logistical services also need their feet on the ground everywhere. Unlike procurement markets and classic outbound logistics, with general cargo and parcels there tend to be few unidirectional routes – networks go to and fro. As a result, all sites – which in network terminology are often referred to as nodes – have to be linked to one another on a daily basis. Each site therefore feeds consignments into the network every day, whenever orders are received from clients. Independent of daily volumes, each site and network connection has to be kept up and running. As a result, the costs for each location and the cost of maintaining links between all sites has nothing to do with actual volumes – they’re fixed. This is actually unlike the variable costs, which are extremely low.
The market for general cargo logistics is dominated by large companies and medium-sized cargo logistics cooperatives. At the moment, the 10 biggest market players have a market share of around 71 percent of the annual turnover of 6.7 billion (source: Kille/Schwemmer (2014), Top 100 der Logistik, p. 119). The general cargo cooperatives are typically formed by between 40 and 60 small to medium-sized logistics providers, who enter alliances to leverage synergies. A small or medium-sized company cannot compete effectively in the market due to the key customer requirement for nationwide coverage, so it has little choice but to join a cooperative. One consequence of the extremely high fixed costs for maintaining storage space and network connections is that competition is extremely price-oriented and margins are extremely low. This makes it all the more important to keep a lid on costs, maintain high standards, or even improve quality, and one of the best ways to do this is to optimize volumes. And optimization means keeping volumes high and uniform. But another problem is that cargo volumes are dictated by external factors (i.e., contracts) and these are awarded to logistics companies on a daily basis (depending on distribution volumes).
One of the particular challenges with this arrangement is that the fixed costs jump up in steps. Costs are fueled by the nodes (distribution hubs) but also by connections between the nodes, for example if extra hands are needed for trans-shipments or extra units are needed for supplying depots. Currently, the daily fluctuations in volumes in the general cargo market can be between 30 and 40 percent. Because the distribution networks are so complex, it’s unclear what causes this volatility. There are also so many clients in the market and these come from such a wide variety of industries that variations in volumes cannot be seen as an inherent feature of the system. Nonetheless, it does appear that the market has recently become even more volatile. Apparently, variations are random and cannot be predicted. This is a major headache for logistics companies, especially given the current situation with shipment variations of up to 40 percent (based on average volumes). Sometimes suppliers only see what volumes will be like just hours before deliveries are due to take place. This leads to a number of severe problems: When planning staffing levels, it’s sometimes not possible to quickly adapt the number of people needed at any given time. So if volumes are higher than average, units are overstretched and quality suffers; if volumes are below average, the personnel costs per consignment are too high.
It’s also almost impossible to adjust technical capacity factors in the short term, such as storage space or vehicles; the problems are similar to those encountered with employees. Some capacity factors cannot be changed at all in the short term and this creates the risk of backlogs. Overall, the fact that volumes are increasingly volatile and it is virtually impossible to predict volumes means that people plan in too much capacity, simply because quality is important and people base estimates on peaks, not troughs. The problem with this is that unused capacity is expensive. Furthermore, if volumes are so low that capacity is underutilized, this weighs heavily on the environment. Although there is a tendency to plan in too much capacity, the increasing volatility has a detrimental impact on standards, since there are more peaks in shipment volumes that exceed available capacities. Quickly buying in additional capacity through the “temp market” is also a short-term solution and this can be particularly expensive.
Understanding the causes of the problems and being able to spot variations early can be a major help in allaying the problem and can thus raise competitiveness. The project team working on the current project have been successful until now in identifying and systematically capturing the causes. They discovered that there are a variety of reasons for variations in volumes. For example, one influence is the weather. In years when the spring is mild and sunny, orders for products like turf lawns (for DIY stores or private customers) are higher than in colder and wetter springs. Aside from public holidays and school vacations, other factors that can affect volumes are special promotions organized by retail chains. That being said, sometimes certain influences overlap, which makes it all the more difficult to identify the exact causes. The aim of the project at TLT Berlin is therefore to work together with the Steinbeis Innovation Center for Research in Transport and Logistics to develop a procedure for forecasting the volatility of logistics volumes and translate this into a forecasting tool.
Prof. Dr. Dirk Lohre is a lecturer at Heilbronn University and is director of FORLOGIC – Forwarding and Logistics Center, a Steinbeis Consulting Center, and a director of the Steinbeis Innovation Center for Research in Transport and Logistics, which are based at Heilbronn University. Both Steinbeis Enterprises offer their clients support with general cargo, contractual logistics, freight forwarding control, and green logistics and sustainability.