Traceability is often broken down into three dimensions:
- Depth – how far upstream and downstream it is possible to trace
- Breadth – the amount of information recorded
- Precision – the ability to pinpoint particular items’ movements
I have tried to illustrate this in the figure above.
Traceability depth means how far upstream/downstream the system can track information. For example, in production of feed for fish farming, the raw material can be traced to finished feed and on to the customer. Traceability depth is increased if the feed can be traced downstream to the cage where the fish lives and to the retailer. Upstream, the traceability depth increases if the fishmeal used in the feed can be traced to the factory where it was produced and further to particular catches taken by specific fishing boats.
Traceability breadth describes the amount of information associated with a product, as it is being moved and processed into other products. Consider again production of fish feed. The breadth of traceability information can grow from covering only the identity and amount of the raw material, to taking into account how it is transported and conditions through the process. Such conditions could be temperatures, humidity and results from analysis. This way the breadth of the traceability is increased – more aspects that may come in handy in understanding the product are available.
Having a wide foundation for traceability is important if the information is to be used for optimization and quality improvement.
Traceability precision decides the accuracy with which a traceable entity’s movement and exposure through a process is recorded. Consider for example industrial production of bread. Several loaves of bread are baked together on large trays. Suppose then that from time to time there is a problem with some loaves but not with others. As it turns out, from the same tray there can both be breads with and without problems! A high precision traceability solution for this bakery would record the position of each bread in the oven, and equip them with identifiers that makes it possible to trace upstream from a customer complaint. Maybe it turns out that the ones with problems were all positioned in the same area as they were baked. Maybe the heat in the oven did not distribute evenly!
I do not think industrial bakeries would have this type of precision, they would probably be satisfied with being able to map from bread type and production day to the raw materials used. And should there be a problem, they are prepared to recall large quantities. However, with more precise traceability information the problem with the oven can be discovered as part of process optimization.
The art of loosing information
The ability to maintain precision in a traceability system is limited by the way production lines operate. For example, if large volumes of raw materials are mixed then precision that existed before that step disappears. So if you were producing tomato ketchup and had farms that supplied tomatoes with high precision, then this disappears if truckloads from different farms are mixed in the production.
Precision is also diluted if products are recirculated under production. Such phenomenons are sometimes referred to as “silos”. Then in theory a particular generation of the production will live on forever in future production. In practice however, such continuous production will be broken into distinct time steps by management routines. But, if you are building a production line from scratch, you may want to spend some time thinking about how to protect investments in precise traceability information.
On the other hand, sometimes it is necessary to loose information in order to be able to make a homogeneous type of product. Consider a factory producing tomato puree. Their customers probably want the “same” product irrespective of where the tomatoes came from. But the tomatoes delivered by farmers may be in varying stages of ripening. So while one batch may be made from perfectly ripened tomatoes, another batch may be based on a blend of unripe and “over-ripe” tomatoes. Hopefully, the producer is so good at this balancing of ingredients that they have the same qualities for all customers. But information about the different farmers is gone.
A different type of traceability
Two times ago I introduced key functions in the GTNet ecosystem. Now I will turn to a special variant of traceability that we call corporate traceability.
The graphic above illustrates how global traceability (lower part) is made up of internal traceability (upper part) for each participant and movement of goods between companies. In the global view, chemicals and grain are delivered to the feed producer, feed is sent to the farm, the farm sends livestock to slaughter and the retailer receives neatly packed pieces of animal protein.
Each of the participants in the lower part are probably independent companies. Therefore they should each have their own GTNet domain, and share global information through GTNet.
But what happens if two or more of the production lines are owned by the same company? In a corporation goods can be shipped both between daughter companies (internally?) and to independent chain partners. Therefore the goods must have unique identifiers. Each corporation can choose between two extremes to create traceability:
- Each daughter company can have its own solution
- The corporation can have a shared solution for all the daughter companies
A TIX can in other words be used both as a complete internal or corporate traceability system.
The illustration above shows these two ways of organizing such a ownership of multiple production lines. Each production line is surrounded by a polygon or rectangle, with trade units moving between them. Grey rectangles are used to indicate production lines owned by the same owner. We call the alternative on the right hand side corporate traceability. In the left hand alternative each production line has its own GTNet domain, whereas in the alternative on the right the production lines with the same owner shares a GTNet domain.
There are a number of issues to consider when deciding which alternative to go for. First of all, company organization and ownership may change over time. So one should bear in mind that what you have today may be different tomorrow. In business life mergers and splits reflect changing business strategies and focus. Aspects like the cost of integration and subscription to GTNet are probably of less importance than other business goals. Anyway, letting each production line have its own traceability model allows each to be specialized, while having one for all production lines in the corporation forces this to cover the union of the involved lines.