Case studies in logistics modelling of pork supply chains

By Willem Rijpkema

European pork supply chains are characterised by a variety of farmer production systems and processing systems. Furthermore, market segments have their own specific preferences for product specifications and logistics services.

 

The complexity of logistics planning in the pork supply chain is further increased by its specific characteristics like: animal welfare, a variety of relevant product quality features (e.g. weight, lean meat ratio), changing quality features over time, and a divergent production process (one pig ends up in a multitude of end products). In recent years, market consolidation and internationalisation have taken place.  Furthermore, the means to gather and communicate quality information for logistics decision making have improved by developments in sensory technology and ICT (Information and Communication Technologies). Despite these recent developments logistics processes are still subject to a number of inefficiencies, like poor material usage, operational inefficiencies or low perceived product quality at final customers. This results in suboptimal market performance and an increase of environmental impact.

 

We try to address these inefficiencies effectively throughout the pork supply chain in a number of case studies. This research, carried out at Wageningen University, is titled ‘logistics modelling of pork supply chains’. Within these case studies innovative logistics management concepts and decision support models are developed and assessed. These models involve multiple actors in the pork supply chain, as well as multiple tasks(supply, demand, distribution, processing).

^{Fig. 1: Schematic overview of pork supply chain actors, their interactions and tasks}

Use of quality information is a key element in effective logistics planning and decision making in pork supply chains. Therefore we develop logistics concepts that exploit (new) product quality information to improve the match between supplied and demanded products. This implies that meat with specific quality features is directed to designated processes and to specific market segments that value these quality features most. As a result the product-market combination as well as the processing performance (less waste, higher efficiencies, etc.) is improved.

 

Slaughterhouse supply planning

In a first case study the use of quality information in slaughterhouse supply planning was analysed. A meat processing company may own multiple slaughterhouses or individual slaughterhouses with different demand for quality features (e.g. weight or lean meat ratio). This is caused by differences in end products which each of the slaughterhouses can produce and end markets they can deliver to. A stochastic programming model was developed to exploit historical delivery data of farmers to estimate the quality of a farmer’s future delivery. This model finds slaughterhouse allocation plans that fulfil demand for multiple quality features at separate slaughterhouses at a demanded success probability or service level, while minimizing transport costs. Our findings suggest that quality-based allocation of slaughterhouses using historical delivery data is a viable technique, and that it provides insight in the trade-off between higher transport costs of livestock and demand for quality features or service level.

 

Water holding capacity

Another Q-porkchains pilot study in which innovative logistics concepts are developed is work package IV.4. In this study Near-InfraRed (NIR) sensor technology is developed for estimation of the water holding capacity (WHC) of meat products. WHC is an important meat quality feature that affects both processing yield and sensory appearance of meat products, as indicated in several Q-PorkChains newsletters. In this pilot study we develop, in co-operation with the VION Food Group (http://www.vionfoodgroup.com/), logistics concepts for product sorting for WHC at processing level using NIR sensor information, and assess these concepts using simulation technologies.

^{Fig 2: Sorting for WHC involves identifying separate markets for different levels of WHC}

 

We develop logistics scenarios for NIR as a sorting tool in ham production and gather relevant process data (e.g. supply and demand data, processing characteristics and volumes). This data is used to model logistics scenarios using simulation techniques to assess advanced quality driven logistics concepts exploiting NIR sensor technology. This will provide advanced insight in the use of NIR technology for sorting of meat products for WHC. Furthermore the study will evaluate the possible gains resulting from its use.

 

In a final case study a decision support model will be developed that assesses the sustainability of international pork supply chain networks. This decision support model will incorporate regional production and consumption data, and can be used to assess the impact of future supply chain scenarios (e.g. higher energy prices, CO2 taxes, legislation, reduced meat consumption) on relevant performance indicators (e.g. cost, energy consumption, product quality, environmental sustainability).

 

By means of our case based studies we aim at deriving generic principles for use of quality information to improve (logistics) control in pork supply chains. In addition, we aim to identify critical success factors for use of advanced quality information in logistics planning.