Recently, it was not so unusual for a computer to take a whole night to calculate newly programmed formulations. However, as a result of the increasing power of computers in the last five to ten years, this whole process of feed formulation has gained momentum, and decision making is now just a matter of seconds. Several developments, like scale-up of companies, quality consciousness, improved technology, and specialisation caused a rapid cultivation of computer software, and service. The scale-up of companies -natural growth and international acquisitions- is said to be necessary to be competitive in an increasingly global market, but therefore caused a change in the use of computers. More people are working on the same projects which means that their actions interact and become more complex. Systems have to be secured to avoid employees having access to all files and being able to manipulate the system. This requires more assistance from the software supplier, and the coaching of system managers in the feedmill becomes a full project to continuously work on.
More locations to produce
Scale-up occurs when companies merge. It is often seen that after a merger, a feedmill has more than one location where the feed is produced. Suddenly the question emerges how and where to use the raw materials in a most efficient and economical way. Logistics become a major part of the management of the feedmills. Multi-plant optimisation is a solution in this, as boundary preconditions change to economic players when, for example, a question could be asked whether or not to use a limited amount of raw materials. A clear overview is made easier with the help of a computer. A feed compounder who starts operating from two different locations, faces a complexity of problems that multiply exponentially.
Within a company, several small business units are often created that have to work together intensively. The working distance between the nutritionists and raw materials purchase increases as companies grow bigger. A nutritionist only thinks of the animal, its performance, and what raw materials to use when formulating their optimal diet. The purchasers of the raw materials are interested in how much they should buy, what the price is, and when it should be delivered, and where most of the time they have to take forecasts of sales and seasonal differences into account. These sometimes conflicting nutritional and purchasing demands can be put together in a computer model that functions as a decision support tool for both. A multi-period optimisation gives the purchaser the global information and the ‘zoom’ functionality gives the nutritionist all the local details. The complexity of modern feedmilling is reflected in the model in Figure 1.
Shrinking markets cause pressure on margins. That is why some companies internationalise by acquiring businesses in more developing countries. In Europe, for example, the knowledge of feed technology is present in the west, where the availability of local raw materials is enormous. However, production capacity and cheap work force is present in Eastern Europe, and so western feed manufacturers expand their business to the east. To be sure that management and workforce in the different countries are speaking about the same subjects, it is important that computer software supports more than one language. In a software program, nutrient values can be recorded once, but the descriptions should be multilingual. To establish such a system, all the information should be collected and controlled in a central place. This avoids the scattering of know-how, and also secures your market position.
Average quality not good enough
Quality assurance is compulsory when offering software to feedmills. New analysing techniques can help to meet set standards. Today an average quality is not good enough anymore, even when taking standard deviations into account. Customers demand a guaranteed quality, or a so called quality with a ‘level of confidence’.
Modern techniques, such as the Near Infra Red Spectrometry (NIRS), make it possible to analyse raw materials per truck or lorry to define their destined silo’s for storage. Current analysing techniques are so simple that at the place of loading, the quality of the raw materials can be checked. This way a feedmiller can already decide at purchasing level where the raw materials should be transported to, or even decide to sell the load, when the quality doesn’t match his standards. A result of this quality policy is that the real quality of the raw materials is better looked at and less emphasis is put on average quality. Taking the varying quality of the different ‘lots’ into account in the formulation system, leads to ‘real’ solutions reflecting real life situations. This will always lead to a better quality of the final product, the feed.
Not only quality control, but the whole process of feed manufacturing has gained from developments in technology. Feed formulation has evolved from batch processing to interactive processing. The mathematical complexity has grown enormously. Two decades ago, a computer needed a whole night to calculate 20 formulas, but now as many as 300 to 800 formulas can be run in just one or two minutes. For example, when a load of soybeans is on offer, the purchaser can put the relevant information of that batch in the optimising program for his feedmill, and whist still on the phone with the seller, tell him immediately if the beans are of interest to him. So at the purchasing level, one can work extremely close to the market.
Back to the future
Modern technology leads us back to where it all started, but this time at a higher level. A few decades ago, one had to log on to a central computer that was in use for everybody who was willing to pay calculation time. The introduction of mini computers and PC’s speeded up the developments as well as accurateness, but became complicated when programs started to live their own lives at different user levels, and the decentralised use of computers sometimes got out of control. Nowadays, when computers and servers become very powerful, it is only a small step to go back to the future and let employees log on to a central computer, which this time is located at the head office. At this central place, the main programs are run and the employee’s computer or laptop only needs start up, and connecting software to communicate with the main computer.
Thanks to the Internet, remote log-on and the extended data security, this line of working has the major advantage that all company’s feed specialists in the field use the same program, only have access to his dedicated part of the software, and cannot take knowledge from the computer when leaving the company. Updates of programs are also made centrally, so everybody always works with the latest software and at reasonable costs.
In the near future, it will be possible to use a cellular phone to connect the laptop to the main computer via Internet. The only barrier in introducing this system to the feed industry is the insufficient capacity of the telephone network, but within a few years it will be fully established. Here multilingualism is again important to make sure that fieldsmen in another country can also work on the central computer at headquarters.
Improved milling processes
Technology also improves manufacturing processes, and makes them more complicated. For example, the dosing of liquids or integrated premix dosing systems have been around for some time. Extruders are comparatively new in animal feed production, and require a complicated control. The raw materials are more processed, which means that the control of humidity is becoming more relevant. Modern formulation software can deal with these problems. To a feedmill, it is very relevant at what moisture levels the feed is sold. A feedmill that produces one million tonnes of feed per year, and sells the feed at a 1% too low moisture level, is missing out on 10,000 tonnes of sales.
When a feedmill is using two different processes to make a compound feed, a multi-level formulation is required. Some feedmillers formulate rations using wet by-products, and one or more concentrates. Other production units powderise wet products, and then mix this powder with other raw materials to form a dry mixture. A multi-level optimisation program is needed here so that one can control all levels, including the production costs of each level, in one single process to get the least cost ‘end’ solution.
The cheapest kilogram of feed does not always lead to the most economical end result. The animals’ performances are influenced by the daily intake of nutrients (energy). By expressing the nutritional requirements per unit of energy (instead of weight) and by entering the raw material requirements as a relation to the weight, the lowest cost price per unit of energy is obtained. This approach provides the cheapest kg. meat, milk, and egg production.
People are becoming more informed through research and modern communication. There is more information on how the animal functions, and on nutritional values of raw materials. Here lies an opportunity for companies that want to specialise. A feed formulation program could link the mathematical model of the animal with nutritional values of feed. Software transforms the nutritional values of the ingredients to optimal performances of the animal.