The computer is a feedlot fixture. It simplifies bookkeeping, spits out payrolls, routes e-mail and runs the Internet. But in the not-too-distant future, the computer will reshape the feeding industry, driving a revolution in beef genetics, feeding practices, the shift to value pricing and the formation of alliances.

Feeders are harnessing the computer's ability to collect, sort and analyze massive amounts of information, allowing them to unravel the complex interaction of hundreds, perhaps thousands, of factors that impact performance.

As this happens, the feeding industry will undergo radical change. Sick cattle will be found and treated faster. Feeders will be able to identify which cattle in a pen perform best, then change buying patterns to get cattle from ranches that produce the best performers. Packers may use computers to determine beef quality with far more accuracy than the current government grading system.

Ranchers, in turn, will use the feedback to identify superior bulls and cows, speeding genetic improvements. Better genetics and grading will speed the move to value-based pricing and the formation of alliances that reward quality.

Data, Data, Data "With the computer, we'll analyze huge amounts of data and modify the way we feed cattle," says Paul Marvin, president of Global Knowledge Group, an Internet development company for the beef industry, in College Station, TX.

Without the computer, such analysis would be impossible. Take, for instance, efforts to analyze cattle performance. It's easy to compile aggregate performance for a pen of cattle but not so easy to analyze individual performance and track genetic backgrounds of 200 or more individual cattle in that pen.

"In the feedlot business, we know very accurately what the pen has done in terms of performance," says Tim Stanton, professor of animal sciences at Colorado State University. "But there's tremendous variability in feed conversion in a pen of cattle on an individual basis. Without knowing what each animal costs us, it's difficult to make huge leaps forward in genetics and cost of gain."

Computers may also help feeders make faster, more accurate decisions. For example, if feed consumption suddenly drops, managers currently might make changes in feeding regimes. In the future, feedyards may be interconnected within a geographical area so managers can quickly determine if some weather phenomenon caused a drop in consumption.

"If you know that the drop in consumption is weather related and it happened in 25 other yards, you're not going to be tweaking your feeding system," says Lee Borck, president of Ward Feedyard, Inc., Larned, KS. "You don't want to tweak the system if there's nothing wrong with it."

A Computerized Feedbunk One major project to analyze feeding behavior is taking place at Cactus Feeders under the direction of Roche Animal Health. The system uses electronic antennae from Growsafe, Inc., of Alberta, Canada, embedded in feed bunks. The antenna picks up the presence of an animal at the feed bunk through an eartag sensor. The antenna checks for the animal's presence every 5.2 seconds and makes an electronic record which shows how often the animal is at the bunk, how long it stays and total time spent at the bunk.

One obvious use is to identify animals off feed, an early indicator of sickness. The system can potentially spot sick animals more quickly than pen riders looking for droopy heads. But the system can do more sophisticated tasks; e.g., matching feeding behavior with weather data to determine if changing weather curtails bunk time.

Another system from Caere Corp., Los Gatos, CA, will allow feeders to track genetic and performance data. It uses a standard numbered cattle eartag, computer database and hand-held optical character reader to scan information and enter it in the database.

With this system, genetic data is recorded on a database by eartag number. Relevant information can be added under the eartag number all the way through finishing. "The tag on the animal is a license plate," says Andrew Allansmith, Caere's marketing manager.

For example, if an animal is treated with an antibiotic, pen workers can scan the eartag to identify the animal, then scan a printed menu to enter the antibiotic dose in the animal's permanent record. This is far easier than carrying a computer keyboard around a feedyard to enter data. Using just one ID number, feedlots create a complete performance, health and genetic record for each animal.

The goal is simple: find what works and duplicate it, or find what doesn't work and avoid it. "There are so many areas that we could be working on," says Ward Feedyard's Lee Borck. "I think the cattle industry is ripe for finding new uses for the computer."

But, as feeders increasingly harness the computer to analyze performance data, they will find themselves buried in data and added expense.

Cactus Feeders has experienced both. It cost roughly $20,000 each to wire just four pens to monitor cattle feeding behavior. And the flow of information has become a deluge.

Cactus Feeders' experience highlights two practical problems that must be solved to make computer analysis a fixture for day-to-day operations. * First, the cost of equipment must come down so feedyards can afford to put it in every pen.

* Second, the systems must be modified so they produce a small amount of immediately useful information rather than a blizzard of research data.

If these hurdles are cleared, feeders can expect to see two types of data analysis projects. One would consist of major ventures in which a small number of pens at selected feedyards would be wired for massive research projects.

At the same time, most feedyards would use less elaborate versions of the same technology to produce a small amount of very useful data. One example would be periodic printouts with the eartag numbers of feeder cattle that aren't eating. Pen riders could pull the animals and check for sickness long before outward signs of illness.

Spencer Swingle, Cactus Feeders' director of nutrition and research, has received a taste of the mountain of information computers can produce. "It becomes a question of how do you filter the data so you get what you need?" he says. "How do you get the data to where you can make management decisions?"

For a while, feeders may have to tolerate high costs and more information than they want. But down the road, they may get the best of both worlds - high-quality research and low-cost systems that can be tailored for widespread use to give them exactly what they need for day-to-day decisions.

The Answers Are In The Questions As feeders, ranchers and meat packers compile more data, they will be able to analyze it in hundreds, thousands, millions of ways. It becomes a matter of choosing which questions to ask the computer.

For example, if you collect data that shows how often a feeder steer goes to the feed bunk, how long he stays, when he feeds and total time he spends at the bunk, you might cross check the data with slaughter weight. It may be that feeder cattle that eat for short periods but eat often gain the most weight.

Or it might turn out that feeder cattle eating for longer periods do best. Nobody knows, but now the feeders can harness the computer to find out.

For example, feeding monitor technology was first used in the ostrich business. At the outset, ostrich chicks had a 10% survival rate. When ostrich farmers turned to the computer to monitor feeding behavior, they could spot an ailing ostrich chick three days earlier. Early detection and treatment boosted survival rate to 90%.

Nobody knows how the flood of data will reshape the cattle business. But the odds are it will help to cut costs, improve genetics and provide quicker, better treatment of sick cattle. And with the chicken and pork industries nipping at its heels, the cattle business can't afford to turn down any help.

Today's feedyards won't survive unless they harness computer technology, but computers won't save bad management.

"Feeders who don't use computers won't make it," says Paul Marvin, president of Global Knowledge Group.

"They'll be able to do it for a little longer, but in four or five years no one will able to do it without a computer."

Marvin says computers are only as good as their operators. If you don't ask the right questions or track the right variables, you can make a wrong decision. Harry Knobbe has a similar view. "We're involved with computers," said Knobbe, owner of a 5,000-head yard in West Point, NE. "But that isn't what makes you money. It's your debt to equity ratio.

"Some people do as well without a computer as those who have one," Knobbe says. "If a computer helps and you can afford it, use it. But don't think it's the difference between success or failure."