Comparison of Software Applications for Formulating Dairy Rations Dan N. Waldner Extension Dairy Specialist Oklahoma State University 

Introduction A variety of software programs are available in the marketplace from both university and private industry. The Nutrient Requirements of Dairy Cattle published by the National Research Council (NRC) provides the foundation for most programs. While this brings some commonality to the software available, generation of new knowledge in dairy nutrition occurs more rapidly than a new NRC can be published. Therefore, developers of software often modify the requirements based on more recently published research, geographical peculiarities or factors not explicitly considered by the most recent NRC. Ration programs will, to some degree, contain similar components or inputs in order to develop a feeding program. These components may include, but are not limited to inputs concerning: description of the animal for which the diet is intended, identification and selection of ingredients, nutrient and ration constraint input, selection of a ration formulation objective, results and analysis of the formulation and report output. However, programs can differ greatly in the amount of input needed, the number of nutrients and feeds used in balancing, the type of objective functions available for ration formulation and completeness and versatility of report writing. These considerations, coupled with the flexibility to tailor the program to the users needs, play a significant role in the decision to purchase and use a particular piece of software. Due to the number and variety of software packages available, selection of an appropriate ration formulation program can be time consuming. The purpose of this paper is to create a guide to aid in software selection by evaluating a variety of commercially available software, from both university and private industry sources. Table 1 contains a list of the software reviewed and the type of format in which they are available. Areas evaluated include: feed composition databases, dry matter intake and nutrient requirement estimates, animal description input options, general data input and management, formulation objective functions, and report writing. While this list is not exhaustive, it should provide the reader with a good jump-start into identifying a suitable ration formulation program.

Agri Visser Consultancy gives a new meaning to least cost formulation: TailoredDiet is freeware.

Feed Composition Databases Knowledge of the composition of feedstuffs is one of the most important aspects in properly formulating diets. Although it is commonly recommended that feeds, especially forages, be analyzed prior to balancing diets, some nutrients or components in forages are frequently not determined, and several concentrate ingredients may not be analyzed even though they can be highly variable in composition. The default values for nutrient concentrations are from tabular values when analytical values are not obtained. The NRC (1989) used various publications to compile feed composition tables, and this publication is commonly used as a source for tabular values. Bucholtz (1997) observed that coefficients of variation for crude protein (CP), neutral detergent fiber (NDF) and crude fat within various commodities ranged from 4 to 35, 10 to 81 and 12 to 60%, respectively. The mineral composition of commodities has been found to be highly variable, often more variable than the organic components. Tables 2 and 3 present the organic constituents for corn silage, ground corn, soybean meal, whole cottonseed, wheat middlings, blood meal and tallow from both the 1989 and 2001 NRC and the various ration formulation programs. For feedstuffs from the new 2001 NRC, net energy of lactation (NEL)values are presented for a cow fed at 3X maintenance. The mean NEL value for all feeds listed in the 2001 NRC are approximately 2% lower than the mean NEL value for the same feeds in the 1989 NRC. Although, on average, the values are similar, some marked differences exist. In general, forages, especially lower quality forages, have lower NEL values, high protein feeds have higher NEL values and starchy concentrates have values similar to those in the 1989 NRC. The NEL for cottonseed is about 16% lower than in the previous edition. The 1989 NRC appeared to be the primary standard used by most of the programs for development of their ingredient databases. Values for CP, NDF, acid detergent fiber (ADF) and fat were similar among the programs. However, the energy values were somewhat variable among feeds from the different databases, which should be expected given that energy is not determined analytically and several methods are used to estimate energy values. 75

Table 1: Ration formulation software evaluated. Company Program Name System Requirements a,b Program Type 

ACS Computer Services 10800 Lincoln Heights Hagerstown, MD 21740 

Dairy Ration System for Windows (DRSW) Windows 3.1/95/98 

Evaluation Least-cost Agri-Data Systems, Inc. 21620 N. 19th Avenue, Suite A-10 Phoenix, AZ 85027 Dairy Ration System II (DRSII) DOS 2.1 or higher Evaluation Least-cost Central Valley Nutritional Associates 3320 E. Mineral King Avenue Visalia, CA 93292 

Formulate2 (Form2) DOS 6.0 or higher Evaluation Multiple objectivesc Cornell Univ. Animal Science Department 130 Morrison Hall Ithaca, NY 14853 

CPM Dairy (CPMD) Windows 95/98/NT Evaluation Least-cost 

Dalex Computer Systems, Inc. 4165 Shoreline Drive, Suite 40 Spring Park, MN 55384 The Consulting Nutritionist (TCN) Windows 95/98/00/NT Evaluation Least-cost 

Easy Systems, Inc. 2550 Northwinds Parkway, Suite 225 Alpharetta, GA 30004 

Brill Formulation Multi-Species Ration Balancer (Brill) Windows 95/98/00/NT Evaluation Least-cost MSU Bulletin Office 10B Agriculture Hall Michigan State University East Lansing, MI 48824-1039 

Spartan Dairy Ration Evaluator/Balancer (Spartan) DOS 2.1 or higher Evaluation Least-cost Profit Source 10458 County Road F Athens, WI 54411 

RationPro (RtnPro) Windows 3.1/95/98/00 Evaluation Least-cost The Ohio State University 221 Animal Science 2029 Fyffe Road Columbus, OH 43210 

Ohio Dairy Ration Program (ODRP) DOS 2.1 or higher Evaluation Least-cost TriLogic Systems P.O. Box 2979 Iowa City, IA 52244 

DairyMax (DMax) DOS 2.1 or higher Evaluation Least-cost Maximum profit Univ. of California Extension Software Support & Distribution Department of Animal Science Davis, CA 95616-8521 

PCDairy-2 (PCD2) DOS 3.3 or higher Evaluation Least-cost Maximum profit aMinimum system requirements. bAll DOS version software is Windows compliant and may be run on Windows 3.1/95/98 systems. cProgram provides for multiple formulation objects. See text.

Agri Visser Consultancy gives a new meaning to least cost formulation: TailoredDiet is freeware.

Significant variation existed among the programs for concentrations of rumen degradable protein (RDP). The 1989 NRC provided limited data regarding the degradability of protein fractions within feedstuffs. Therefore, most of the software programs have undoubtedly relied on other sources for this information. Dry Matter Intake and Nutrient Requirements Establishment of dry matter intake (DMI) and nutrient requirements is another important step towards developing a properly formulated ration. The animals used for the comparisons in this paper are described in Table 4. It was assumed all cows were housed in a thermoneutral environment, in at least their third lactation and had a constant body weight. These assumptions were made to minimize confounding of how the programs determine nutrient requirements. The suggested DMI and nutrient requirements from the different sources were taken directly from the computer programs, including the diskette provided with the 1989 NRC. There was no distinction in estimated DMI or nutrient requirements for dry cows between the 1989 NRC and the programs PCD2 or DMax (Table 5). Dry matter intake predictions for dry cows greater than 21d and 21 to 0d prepartum ranged from 1.45 to 2.05 and 1.43 to 2.0% of body weight, respectively. Six programs (PCD2, DMax, DRSII, DRSW, RtnPro and TCN) did not reduce their DMI predictions for cows within 21d of calving, which is similar to the 1989 NRC. The 2001 NRC reduced DMI 8% from 1.97% of BW for far-off dry to 1.82% of BW for close-up dry cows. The most significant change in DMI prediction came from the ODRP, Spartan, Brill and CPMD with reductions of 12.3, 17.5, 18.0 and 20%, respectively. Estimates of CP intake for far-off and close-up dry cows ranged from 2.31 to 3.67 and 2.31 to 3.29 lbs/d. Crude protein requirements were similar for the 1989 and 2001 NRC, 2.85 and 2.82 lbs/d, respectively. Five programs had similar values, which included ODRP, PCD2, Brill, DMax and DRSW. Spartan had the highest CP intake estimate at 3.67 lbs/d. Net energy of lactation requirement estimates for both far-off and close-up dry cows ranged from 13.0 to 15.6 Mcal/d. Net energy of lactation requirement estimations from the 1989 and 2001 NRC were 13.5 and 13.0 Mcal/d, respectively. Eight programs (ODRP, PCD2, Spartan, Brill, DMax, DRSW, Form2 and TCN) had values similar to the 1989 NRC, while CPMD was in close agreement with the 2001 NRC value. Calcium and phosphorus requirements for dry cows were similar among all of the programs evaluated. Estimated DMI for lactating cows at 30, 60, 100 and 250 DIM averaged 3.18, 4.15, 3.78 and 3.02% of BW, respectively (Table 6 and 7). There was no difference in estimated DMI or nutrient requirements between the 1989 NRC and DMax across all stages of production. Form2 was similar to 1989 NRC at 30 and 60 DIM, but not at 100 and 250 DIM. DRSII was similar to 1989 NRC at 30, 60 and 100 DIM, but not at 250 DIM. Estimated DMI was highest for Brill at 30 and 60 DIM, 3.60 and 4.32% of BW, respectively and highest for 2001 NRC at 100 and 250 DIM, 4.10 and 3.31% of BW, respectively. A recent comparison among 1989 NRC, ODRP, CPMD and Spartan (Eastridge et al., 1998) for predicting DMI in early lactation (≤ 8 weeks), showed that overall the 1989 NRC provided the most accurate estimation of DMI. However, at ≤ 4 weeks of lactation Spartan provided the best estimate of DMI (bias = + 0.44 lbs/d) among the programs evaluated in that study. For the current paper, the estimated DMI for cows 30 DIM from Spartan was 3.11% of BW. Dry matter intake estimates lower than Spartan were witnessed with the 2001 NRC, CPMD, RtnPro and TCN programs at 2.97, 3.01, 2.73 and 2.92% of BW, respectively. According to the discussion, the 2001 NRC DMI equation predicts DMI very closely to actual DMI for the first 10 weeks of lactation and then slightly under predicts DMI thereafter. This is interesting since the 2001 NRC produced the highest estimates of DMI at 100 and 250 DIM in the current comparison, indicating that all of the programs may be significantly under predicting DMI of cows greater than 100 DIM. Estimated CP intake requirements for lactating cows at 30, 60, 100 and 250 DIM averaged 7.89, 10.85, 9.37 and 6.55 lbs/d, respectively and consisted of considerably more variation than estimates of NEL requirements. Variation among the programs was least for predictions of CP intake at 60 DIM and greatest at 250 DIM. Among the programs evaluated only DMax had protein values in agreement with the 1989 NRC for all stages of production tested. The ODRP had the highest CP intake requirement at 30, 60 and 100 DIM, (8.64, 11.4 and 10.1 lbs/d, respectively), while DRSII had the highest requirement at 250 DIM (7.43 lbs/d). The lowest CP intake requirement predictions for 30, 60, 100 and 250 DIM were 7.58 (TCN), 10.7 (Spartan), 9.18 (1989 NRC, DMax, DRSII and Form2) and 6.04 (TCN) lbs/d, respectively. As indicated previously, NEL requirement values contained significantly less variation than protein requirement values for the lactating cow examples. Net energy of lactation values averaged 35.7, 47.2, 41.5 and 29.9 Mcal/d at 30, 60, 100 and

250 DIM, respectively. Four programs, PCD2, Spartan, DMax and TCN, had identical energy values when compared to the 1989 NRC across all stages of production examined. The ODRP provided the lowest NEL estimates. Values from the ODRP at 30, 60, 100 and 250 DIM were 35.3, 46.1, 40.5 and 29.0 Mcal/d, respectively. Net energy of lactation values from the 2001 NRC were 36.0 (30 DIM), 47.1 (60 DIM), 41.7 (100 DIM) and 29.6 (250 DIM) Mcal/d. The 36 Mcal/d estimate from the 2001 NRC at 30 DIM was the highest among the programs for that stage of production and was similar to the CPMD estimate. Dry matter intake is the foundation from which rations are balanced; therefore, the accuracy of DMI prediction or estimation is extremely important. The most common factors used in the prediction of DMI are BW and FCM yield. Other factors incorporated into some equations include stage of lactation, parity, milk component yield, dietary components and environmental variables. Roseler et al. (1997) concluded that milk yield, BW, diet and management, climate and body condition score accounted for 45, 17, 22, 10 and 6%, respectively, of the variation in DMI. Milk yield and BW accounted for 62% of the variation, which explains why these variables are the most common in equations for predicting DMI. However, it should be remembered that no prediction equation is as accurate as actual measurement of DMI on the farm. Indeed, the 2001 NRC committee reminds its readers of this fact. All of the programs evaluated allow the user to enter their own DMI values directly, thus circumventing the use of the DMI prediction equation. Further, some programs allow the user to edit the prediction equations to more accurately reflect current research or field observation. Overall, differences existed in recommendations for nutrient intakes among the programs. These differences in recommended nutrient intakes by the programs were accentuated by differences in DMI predictions when the nutrient recommendations were compared based on dietary concentration. Further, several dietary components existed in the computer software programs for monitoring purposes that do not exist in the 1989 NRC. These additions reflect changes in the field of dairy nutrition since the 1989 NRC recommendations and the freedom of software developers to include variables for which limited data may exist. Animal Description Table 8 contains a list of the animal description and production input information items for lactating and dry cows of the different computer programs. Animal descriptors required to generate lactating cow requirements from the 1989 NRC program diskette include: body weight, milk production, milk fat percent, days pregnant, lactation number, live weight change and a lead or discount factor for feed or NRC assumed energy concentration. The 2001 NRC allows input for up to seventeen items to describe the lactating cow and up to eleven input items for description of the dry cow. Only milk lactose percent, required by the 2001 NRC, was not part of the input needed by any of the other programs evaluated. All of the programs evaluated required an estimate of live body weight, milk production and milk fat percent for determination of lactating cow DMI and nutrient requirements. Other popular input items included, live weight change (11 programs), lactation number (9 programs), breed or breed type (5 programs) and milk protein percent (5 programs). The program most similar to the 2001 NRC for animal input information was CPMD. However, it should be noted that TCN is also similar to the 2001 NRC, when using the modified Cornell model portion of the program. Further, the TCN program allows for selection of one of three amino acid requirement calculation schemes and the use of one of two different DMI intake equations. Information required by both the 2001 NRC and CPMD but not the other programs included: age, age at first calving, body condition, calf birth weight, calving interval, and mature weight. Among the programs evaluated, four (01NRC, CPMD, PCD2 and Form2) allowed for an activity or grazing allowance, three (PCD2, DMax and Form2) allowed for input pertaining to the percent of 1st and 2nd lactation animal in the group and only Brill and DMax provided the ability to use a lead factor for lactating and lactating and dry cows, respectively. Six programs allowed for input of various environmental and housing conditions for both lactating and dry cows, however the effect that these input items had on DMI and (or) nutrient requirement calculations were not readily apparent for all of the programs requesting this input. Again, several items used to define the animal for which a particular ration is being formulated varied greatly among the programs evaluated and included items not considered by the 1989 NRC. Data Management and Ration Formulation Table 9 compares the major data management features of the different computer programs. The 2001 NRC program was not considered for this discussion since it is an evaluation program only and not intended for direct commercial use. In consideration of the major design and setup options among the programs evaluated, ODRP and DRSW did not allow for use of both the imperial and metric systems of measurement and ODRP did not allow interchangeable moisture 82

Agri Visser Consultancy gives a new meaning to least cost formulation: TailoredDiet is freeware.

basis input selection. The programs CPMD, ODRP and RtnPro restrict the user to input of ingredient weights and ingredient usage constraints on an amount per head basis only. All of the programs allow for editing of the nutrient content of ingredients contained within the programs libraries. The maximum number of nutrients allowed by the various programs ranged from 29 (ODRP) to 200 (Brill and TCN). The nutrient counts for CPDM (60), ODRP (29), Form2 (48) and RtnPro (38) are static and do not allow for the creation of new nutrients by the user. Brill, DMax, DRSII, and TCN offer the user the opportunity to edit or write new formulas for calculation of existing or user created nutrients. With the use of multiple libraries all of the programs except ODRP have no practical limit on the number of ingredients that can be edited, created and stored by the user. Further, all of the programs allow mixing of multiple ingredients to form a new ingredient, and all programs except Spartan, allow the user to easily save the newly formed ingredient in an ingredient library. Only CPMD and TCN had no limit on the number of ingredients allowed in a single formulation run. The ingredient use restriction per formulation for the other programs were ODRP (20), PCD2 (30), Spartan (79), Brill (40), DMax (50), DRSII (40), DRSW (30), Form2 (40) and RtnPro (50). All of the programs reviewed can function as a ration evaluator and allow the user to develop a ration formulation at least-cost. Both the DMax and PCD2 programs have a profit-maximizing objective function, while only Form2 offers up to four different optimization functions. These options include: a least cost function, minimization of concentrate costs, minimization of forage costs or minimization of user-tagged feeds. Six programs (Spartan, Brill, DMax, DRSII, DRSW and TCN) allow for the creation of new nutrient constraints and all but Spartan allow nutrient constraint formula editing. Along with nutrient constraints, the use of group and ratio constraints can be extremely helpful in controlling formulation results. All of the programs except CPMR, ODRP, and Spartan allow the use of such constraints. The CPMD and ODRP are the only programs that do not balance for minerals (except Ca and P) or vitamins and only PCD2, Spartan, DMax or RtnPro do not provide amino acid analysis. An additional option within the Brill, DRSII, RtnPro and TCN programs is the ability to formulate multiple rations at the same time. While Form2 will allow a similar process, using a common feed library, it requires many more steps than the previously mentioned programs. Troubleshooting infeasible solutions can be a frustrating task. The use of dummy ingredients in a solution provides useful feedback to the user and can aid in determining additional ingredient or nutrient needs to bring about a reasonable solution. Programs incorporating this type of feedback include CPMD, PCD2, Spartan, Brill, DMax, and DRSII. Although TCN, RtnPro and DRSW do not allow for use of dummy ingredients, these programs will return solutions with flagged or highlighted nutrients indicating where problems exist in returning the infeasible solution. The Form2 program uses a text message box to indicate an infeasible ration was returned and limiting nutrients. Some of programs incorporate the use of a text message in combination with dummy ingredients or flagged nutrients to aid the use in identifying problems and possible solutions. As mentioned previously, flexibility is an important consideration when deciding to purchase a particular software program. Overall, it appears that the commercial software products allow for more flexibility and user input than the university developed software. Among the commercial programs available, RtnPro appears to provide the fewest options for managing data and generating formulations while, Brill, TCN, DMax and DRSII provide the most flexibility. Print Reports Table 10 consists of the print report features and options from the different computer programs. The kinds of output from the programs are as varied as the programs themselves. At a minimum the user should be able to produce a concise comprehensive ration report containing a title describing the output, a description of the animal on which the evaluation was done, a summary of the type and amount of ingredients used, nutrient content of the ration and cost. This minimum requirement; however, is hardly enough to satisfy the needs of most users. Again, the university-developed programs offer fewer options in report writing than the commercial products. Both the ODRP and DMax programs provide little in the form of a comprehensive report previously described. Among the commercial programs, DRSII provides the most comprehensive list of reporting options. Batch or load mixing weights and ingredient blend weights could be provided by all of the commercial programs evaluated. However, there was considerable variation as to the organization and options available within these reporting sections. 86

Surprisingly, several commercial programs (DRSW, Form2, RtnPro and TCN) did not provide for reporting of shadow pricing for ingredients used. Moreover, neither Form2 nor TCN provided opportunity pricing for ingredients not used. This information can be important in making commodity-purchasing decisions and should be included as part of the reports package and not just for on-screen display. Summary Because a variety of software applications for formulating dairy rations are available, selection of an appropriate program can be frustrating and time consuming. As with any type of software search, the user should begin assessing software needs by listing the important program functions and capabilities desired. All DOS software versions reviewed are highly compatible within the Windows environment and most of the companies have indicated that Windows versions of their products were either under development or scheduled for release. The 1989 NRC provides the foundation for most programs and brings some commonality to the software available. All software programs incorporate some type of ration evaluation option within their software package. After that the programs vary in complexity. Most software packages provide for formulation of rations at least cost; however, a few also contain other objective functions. These may include a profit-maximizing function as well as functions to allow for optimization of the use of certain ingredients, such as homegrown feeds, while minimizing purchased feed costs. Still other items such as the report writing, economic analysis, diagnosis of infeasible solutions, and overall flexibility of the program are important and vary considerably among the programs evaluated. Overall, private industry software products allow for more flexibility and user input than the university developed software. Further, although more expensive, private companies offer services not provided by universities such as software support and training. 

References Bucholtz, H.F. 1997. Variability in chemical composition of commodities. Pages 113-123 in Proc. Tri-State Dairy Nutr. Conf. M.L. Eastridge, ed. Ft. Wayne, IN. The Ohio State Univ., Columbus. Eastridge, M.L., H.F. Bucholtz, A.L. Slater and C.S. Hall. 1998. Nutrient requirements for dairy cattle of the National Research Council versus some commonly used ration software. J. Dairy Sci. 81:3059. National Research Council. 1989. Nutrient Requirements of Dairy Cattle. 7th Ed. National Academy Press, Washington, D.C. National Research Council. 2001. Nutrient Requirements of Dairy Cattle. 7th Ed. National Academy Press, Washington, D.C. Roseler, D.K., D.G. Fox, L.E. Chase, A.N. Pell and W.C. Stone. 1997. Development and evaluation of equations for prediction of feed intake for lactating Holstein dairy cows. J. Dairy Sci. 80:878. 88

********************

Agri Visser Consultancy gives a new meaning to least cost formulation: TailoredDiet is freeware.

Aqua Feeds: Formulation & Beyond recently conducted a survey of companies that develop software

solutions for feed formulation. Ten companies that develop feed formulation were contacted and seven

companies responded to the survey questionnaire. The questionnaire had several open-ended questions on

the products and services offered by the companies and their target markets. The survey results reveal

companies offering a diverse range of products. Feed companies considering to upgrade or buy formulation

software need to take into account several factors in their decision. They are:

The need depends on their size, number of feed plants and products that they have, complexity of the products, and whether they intend to integrate the formulation software with other applications, for example, lab data systems, bag label generation, ingredient inventory systems, etc. Special needs such as traceability must be considered as well. The ingredient market in which they operate is also an important factor. A company that is planning on using only corn and soybean meal may accomplish their formulation objective with a simple spreadsheet (see box on UNEform). Formulation software generates formulas based on the data provided by the feed company. The primary data input in any formulation system are the ingredient composition data (such as the proximate composition values typically obtained from the lab) which drive various computational models to derive other nutrient values such as amino acid composition.

The feed company should have access to reliable and accurate input data. It should also ideally have a system in place to update the data periodically and dynamically. If the company does not have a reliable data source, it will have to find out from the software providers whether such data can be procured. Computer hardware and software evolve everyday and compatibilities change over time. A formulation company that offers products across multiple platforms is preferable to one that is limited to one platform. More importantly, the company should be constantly upgrading its products to take advantage of new capabilities offered by developments in computer science. Most formulation software are designed to run on a standard, store-bought desktop or laptop. Only when highly complex problems such as large multi-blends with multiple price sets; multi-plant optimizations and associated logistics are to be solved, an advanced computer system will be required.

Post-sales support services offered by the formulation company.

Agri Visser Consultancy gives a new meaning to least cost formulation: TailoredDiet is freeware.

Integration capacity with other company software, especially manufacturing and accounting.

This includes personnel training, troubleshooting, customization, help desk, etc. It will be diligent to find out whether the formulation company has on its staff sufficient resources to handle support in relation to its customer base. This will be especially important in countries and companies that have limited IT resources. Formulation is only one step in the feed business that converts raw materials into finished feed. Capacity and ease of integration of the software is essential if formula changes need to be implemented rapidly throughout the organization. Following is a brief profile of the seven feed formulation companies that participated in our survey. 

ADIFO NV Location: Belgium Year of Start: 1974 The company offers BESTMIX, a least-cost formulation and optimization package that is Microsoft-certified. The software contains powerful mathematical models that support the user with MRP (material requirement planning) by linking product development, purchase, production and sales. The software allows easy expansion and internationalization for companies that have several production units and offices. It is available in several languages and offers options to implement diverse quality standards (ISO, HACCP, GMP, etc.) and traceability. It also provides an optimal transfer of knowledge in the Supply Chain and in an E-business environment, enabling feed manufacturers to build their relationship with suppliers and buyers. 

Creative Formulation Concepts, LLC Domit & Domit Ltd. Feed Management Systems (Formerly known as Agri-Data Systems) Location: Maryland, USA Year of Start: 1978 The company offers software solutions in two Windows based versions called Concept5 and Concept4. Concept5 co Concept5 provides flexible configuration choices ranging from a single user to an unlimited number of concurrent users and manufacturing plants. The Concept4 systems are limited to single users and plants and offer basic least cost features.

Location: Parana, Brazil Year of Start: 1986 The company's product, OptiMix is offered as a complete software package to create and manage feed formulas for small to large feed producers. Easy to learn and use, with embedded multi-client, multi-plant, multi-product, and multi-user capabilities, the product offers a user-friendly Windows interface and allows data to be exported to MS Word and Excel. The Corporate package can work in a network and can be integrated with database management software like Oracle, MS SQL, etc. It also provides a tool for data integration between a main office computer and remote, off-line systems. The company enables on-line support via the Internet.

Location: Minnesota, USA Year of Start: 1968 The company's flagship product is Brill Formulation, an optimization application that is designed to handle formula creation, large batch optimization problems as well as multiproduct, multi-plant, multi-price optimization. The company offers software solutions for generating feed tags, pricing and quoting, and feed ration balancing (for in-field customer ration development), and a real-time information management system that ensures correct formulas are being used in production with complete trace and track capabilities. Contains a broad range of standard formulation functions and is the required database for integration of advanced formulation and management options. Modules such as the multi-step model; multiple product optimization for one or more plants; product costing/pricing, labeling and purchasing can be added at any time. 

UNEForm: A Free Feed Formulation Software Microsoft Excel offers linear programming capabilities that are sufficient for teachers, researchers and students who want to teach or learn feed formulation or create simple, experimental diets. The tool is called “Solver” and is offered as an Add-In. Dr. Evan Thomson at the Animal Science department of the University of New England, Australia has developed an MS Excel spreadsheet file called UNEForm.xls that is already configured to formulate animal feeds. The file can be downloaded at http://ansc.une.edu.au/ansc/nutrition/downloads.html. 

Feedsoft Corporation Location: Texas, USA Year of Start: 2004 The company offers two versions of its formulation software, Feedsoft Formulation. The Standard Edition is an affordable entry level system designed for animal practitioners who want to easily create least-cost feed formulations from a single database composed by any number of available ingredients. The advanced Professional Edition is the tool for those who work with multiple ingredient databases or need to manage ingredient data and formulas from multiple clients and sites. The Professional Edition includes additional functionalities such as multi-database support, price manager, batch formulation, and an import/export wizard.

Flanders Nutritional Technologies N.V. Location: Belgium Year of Start: 1999 The company offers its formulation software Alfablend in two versions. The Professional version is a single product optimization with unlimited number of raw materials, nutrients, and recipes. It also offers tools for labeling and liquid feeding. The Enterprise version contains all the features of the Professional version, plus multi-product optimization across multiple plants. It also contains an easy-to-use tool for premix extraction out of a complete recipe. The software works on a native Windows platform and enables exporting of all information to pre-formatted MS Word templates for printing and documentation purposes.

Format International Limited Location: Surrey, UK & Missouri, USA Year of Start: 1981 ® Single-Mix maintains the ingredient and product database used in all optimization programs developed by Format. The product has the ability to generate key pricing and management reports and offers a range of additional tools such as parametrics. Advanced products from Format handle multi-product, multiplant and multi-time period optimization. It also offers a multicomponent formulation/recipe optimization system that allows the design and construction of products that include subcomponents each of which carries its own specification. Other advanced capabilities include quality assurance checks on formulas, product labeling, full formulation data archiving and interfacing options. The software is designed for Windows operating systems in standalone, client-server or terminal server configurations. Contact person: Ian Mealey

**********************

PCDAIRY-2, a package of computer programs for formulating or analyzing rations for dairy cattle, has been developed with animal nutrient requirements and feed composition data contained in the National Research Council Bulletin, Nutrient Requirements of Dairy Cattle, Sixth Revised Edition, Update 1989. PCDAIRY-2 has been designed to run on IBM PC compatible computers with 512K of memory. This paper describes some of the programs in the PCDAIRY-2 package. Examples of the data input screens and ration printouts will be shown. 

SUMMARY OF PROGRAMS

PCDAIRY-2 consists of eight different programs:

MAXIMIZE weight, milk yield, fat test, weight change, feed prices and milk price.

LC change, stage of lactation, and feed prices.

GROWING  feed prices.

ANLSIS-L National Research Council nutrient standards, and lists the amount of milk that is possible from the ration and the limiting nutrients.

ANLSIS-G cNational Research Council nutrient standards, and lists the amount of weight gain that is possible from the ration and the limiting nutrients. 

FEEDLIST a

DELIVERY cows, and for unloading the mixed ration to a specified number of cows.

Department of Animal Science, University of California, Davis, CA 95616

FEEDTAG calculates the energy content of feeds based on their proximate analysis.

PCDAIRY-2 was designed to minimize the amount of typing necessary to enter feeds or evaluate a ration. Popup menus appear with a list of the choices for entry. The arrow keys on the computer can then be used to select the entry. Feeds are stored in a standard feed library (from the NRC Nutrient Requirements of Dairy Cattle, update 1989) and can be selected to enter directly into the ration. Input and output can be expressed in either English units (lb) or in metric units (kg). All of PCDAIRY's programs incorporate state-of-the-art microcomputer software, including full-screen editing for input, pop-up help message windows, and extensive error checking.

Users may select either least cost or maximum profit basis for ration formulation. Figure 1 shows the screen used to input data for formulating a ration for a 1400-lb cow using the LC program. Figure 1. Cow information: LC program. LC and GROWING program users may select the type of protein to use as a minimum constraint when the ration is formulated: crude protein, undegradable and degradable intake protein, or all three. Similarly, users of the LC, MAXIMIZE, and GROWING programs may select the fiber formulation basis: effective crude fiber, acid detergent fiber, neutral detergent fiber, or combinations of these. In addition, GROWING program users may select the type of energy system to use for ration formulation: the net energy system (NEM and NEG) or the total digestible nutrients (TDN) system. PCDAIRY-2 then calculates the NRC nutrient requirements for the ration, based upon the cow information (National Research Council, 1989). The program allows users to override any of the default NRC nutrient requirements prior to ration formulation. 3 After entering the cow information, users select the feed ingredients available from the standard (STD) feed library and their prices ($ per ton, $ per cwt, or $ per kg). Ingredients not included in the standard library may be entered into the program, with the users providing the required nutrient data. Users may specify a minimum constraint on a feed by entering a constraint amount and then selecting the proper constraint type from the following list of choices that appear in a pop-up window: LBAF (KGAF) = lbs (kgs) as fed LBDM (KGDM) = lbs (kgs) of dry matter %CDM = percent of concentrate portion of ration on 100% DM basis %RDM = percent of roughage portion of ration on 100% DM basis %TDM = percent of total ration on 100% DM basis %CAF = percent of concentrate portion of ration on an "as fed" basis %RAF = percent of roughage portion of ration on an "as fed" basis %TAF = percent of total ration on an "as fed" basis In addition, users may specify a maximum constraint on a feed in the same manner. The concentrate ingredients provided by the standard feed library have default maximum values assigned to them. They may be changed in the manner described above. Figure 2 shows the feeds available for a sample ration using the LC program. Figure 2. Available feeds: LC program. Users may customize the ration formulation by placing minimum and maximum constraints upon groups of feeds, such as wheat products. Individual feed, feed group, and nutrient ratio constraints may also be specified. 

THE FORMULATION PROCESS: USING INFEASIBLE FEEDS

During the ration formulation process, PCDAIRY-2 automatically appends a set of "infeasible feeds" to the feed list. These are "fake" feeds which PCDAIRY-2 will use to avoid getting an infeasible solution when it cannot fulfill the nutrient requirements with the available feed ingredients. The names of the infeasible feeds are descriptive of the nutrient deficiency or excess in the ration. All infeasible feeds will enter the ration only if 4 there is no other way to fulfill the nutrient requirements within existing constraints. When infeasible feeds appear in the formulated ration, users should either modify the existing constraints or add ingredients that can be used to satisfy the nutrient deficiency. 

RATION FORMULATION PRINTOUT

The following information is included in the ration printouts: 1) Ration composition 2) Price ranges for feeds used in the ration 3) Opportunity prices for feeds not used in the ration 4) Nutrient analysis of the ration 5) Nutrient composition of all available feeds Figure 3. Ration composition: LC program. 5 Figure 3 shows the composition of the least cost ration formulated using the data from Figures 1 and 2. Price ranges for feeds used and opportunity prices for feeds not selected are shown in Figure 4. Figure 5 shows a portion of the estimated nutrient analysis of the least cost ration. Minimum and maximum nutrient constraints that were in effect at the time of formulation are listed as well as minimum and maximum constraints suggested in the NRC dairy bulletin (National Research Council, 1989). Figure 4. Price range and feeds not used: LC program. Figure 5. Nutrient analysis of ration: LC program. 6 After formulating a ration, users have the option of combining feeds into a mix. Only real (feasible) feeds that were used in the ration are eligible to be combined. Users must specify a new feed number and name for the mix, and the amount of each feed to be combined. PCDAIRY-2 then adds this new mix feed to the list of available feeds. This option is commonly used to create a concentrate mix. Figure 6. Loading schedule: DELIVERY program.  

THE DELIVERY PROGRAM

The DELIVERY program calculates feed loading and unloading schedules (up to twelve) for any number of animals. DELIVERY produces a loading schedule that lists the order in which feeds are to be added, the amount of each feed, and the scale reading after each feed is added to the mix. If a feed is to be fed separately from the mix, an amount per feeding will be listed. The unloading schedule lists the amount of the mix that is to be fed per feeding for the number of animals specified. Figure 6 shows the loading schedule for the ration formulated using the LC program. 

REFERENCE National Research Council. 1988. Nutrient requirements of dairy cattle, sixth revised edition (update 1989). National Academy Press, Washington, D.C.

************

Format International

Agri Visser Consultancy gives a new meaning to least cost formulation: TailoredDiet is freeware.

Multi-Mix® optimises ingredient use for multiple products over a given production period taking into account quantities of available ingredients. It consistently reduces the total Raw Material cost for a single period production schedule.

Multi-Mix® can be used to plan production, ingredient purchasing and trading. New algorithms take account of buying, stock and selling prices and the available quantities of each category. Scarce and surplus ingredients are allocated to achieve the minimum overall cost. Trading opportunities can also be investigated, balancing potential profits and losses on sales against the value of ingredients when used in production.

Multi-Mix® features – •

1. Simultaneously considers up to three sets of price and availability levels for each raw material; buying, stock and sales.
2. • Advises whether to buy, sell or use ingredients, and reports trading profits/losses from sales.
3. • One-touch, interactive editing and re-optimisation.
4. • Detailed difference reports for comparison of the effects of changes using two solutions stored; current and previous.
5. • Parametric option explores of alternative solutions and price changes on any market cost, purchase cost or selling price for any ingredient.
6. • Fixed formulas may be specified so that ingredients required to make them are included in the overall allocation and purchasing decisions.

******************************

True multi-level optimisation module p

Agri Visser Consultancy gives a new meaning to least cost formulation: TailoredDiet is freeware.roduct information sheet

Today’s advanced Petfood recipes may contain many sub-components or kibbles, each of which carry their own specification in terms of "Claims" and other characteristics. The attributes of the complete product are also defined in terms of nutrition, moisture, ingredients and many other parameters. The challenge is to specify, and to optimise, the whole product simultaneously with all its sub-components.

Integra-Mix^{™ does just that. It is a multi-component optimisation package which formulates multiple sub-components both within products and when they are shared across products.}

Agri Visser Consultancy gives a new meaning to least cost formulation: TailoredDiet is freeware.

*******************

New Century™ software provides a user-friendly, efficient tool for the optimisation and evaluation of recipes and ingredients, with proven cost-saving results. An additional tool available in the New Century™ range of programs to help with evaluation is Parametrics.

By allowing factors involved in the optimisation of recipes to be altered, Parametrics allow the effect of these changes to be quickly examined. It can therefore present a significant improvement in the speed of evaluation and in the quality of information obtained.

Parametrics are available as a separate package for use with single product optimisations in Single-Mix®. This allows the following factors to modified in a recipe optimisation –

1. Raw Material cost
2. • Raw Material inclusion level
3. • Raw Material nutrient level
4. • Nutrient level

By modifiying the value of each factor in a step-wise manner, the effect of this gradual change on product cost, nutrient analysis and raw material use may be quickly investigated.

Parametrics are also available as standard within the Multi-Mix® module, allowing the modification of –

1. Raw Material costs in the Multi-Mix® solution

•

Multi-Mix parametrics are therefore a powerful tool in the evaluation of Raw Material price movements and their effect on the use of other ingredients.

***********************

Agri Visser Consultancy gives a new meaning to least cost formulation: TailoredDiet is freeware.

The Spartan Software Laboratory is part of the Department of Animal Science at Michigan State University. We are responsible for the creation and support of nutritional software programs for livestock. We are currently working on a Windows version of our popular SPARTAN Dairy Ration Evaluator/Balancer. Additional information about this future product will be posted here once we have a program available for beta testing.

The DOS versions of our ration evaluator products are no longer available for purchase. However, the current DOS versions of the Dairy and Beef ration evaluators can be downloaded for free from this website. Unfortunately, the DOS versions of our older programs for balancing horse and swine rations are no longer available.

All of these programs are used to properly balance feed rations so that the animals receive the proper nutrients, in the proper amounts, based on the animals' condition and use. The first three chapters of the user manual for the Dairy program is also available for download in pdf format from the Products section.

Site Overview:

* The Products section gives specific information about each of these products, including screen shots from the individual programs.
* The Customer Support section contains FAQs to assist you with trouble shooting and lists the e-mail addresses of the people you can contact for technical or nutritional difficulties you are experiencing with your SPARTAN program.
* The Personnel section gives a brief introduction to the MSU faculty involved with the SPARTAN project.
* The About Us section describes the mission and goals of the Spartan Software Laboratory.