Computer-Aided Estimating - A Brewing Revolution

With computer-aided estimating systems today, the user has a wide choice - capability wise. The first choice to be made is between accuracy and economy. Many of the "economy" choices are made without concern for accuracy issues. For the "accuracy" issues, however, there is no source where companies can discover what those choices involve. The important fact to remember when looking at computer-aided estimating systems is that each time you send a quote to a customer, you are betting your business on the accuracy of the times in the estimate.

Shop owners and manufacturers continually ask if it really is important to understand the strategies computers use to "build" an estimate. It is. If you are considering estimating software, this knowledge is absolutely necessary. The industry has neither a standard definition for an estimate, nor methods to compare the results provided by "estimating" programs.

In just one year, the available, affordable technology has advanced at amazing speeds. However, before looking at where computer-aided estimating is headed, it's good to review the often misunderstood term "estimating."

What Is Computer-Aided Estimating?

A better question might be: When is a computer-generated price really an estimate? What do you need to know when people talk about estimating software? Typically they refer to one of two groups of software:

• Quoting software simply helps determine a price. The estimator manually calculates or guesses the time it will take.
• Estimating software provides not only a quote, but it also uses feeds and speeds to estimate the actual times that the job will require.

Time is the factor that makes it simple to differentiate between quoting and estimating software. The important fact to remember when buying estimating software is that each time you send a quote to a customer, you are betting your business - profit or loss - on the accuracy of the times in the estimate.

Within the estimating classification, some software systems are standards-based, some are engineering-based, some exhibit genuine intelligence by emulating actual machine motions and others are evolving from one form toward another. Intelligent simulation attempts to determine how long each operation really will take as opposed to how long it should take. And this year, automated feature recognition (see table at the end of this article.) has entered the picture.

Having the capability to account for issues like rapid travel and idle times defines the difference between the standards-based, engineering-based and intelligent simulation estimating systems. The latter system tells you that it will take one minute - not that it should. Intelligently simulated estimates attempt to include every significant event, as opposed to what occurs when interpolating averages times and indicating what should happen.

The term "estimating" still sounds like a guess, no matter how sophisticated the software is. In fact, some computer-aided estimating software programs are true engineering-based process planning and profit-predicting systems. Today, however, just as speed and accuracy are demanded on the shop floor, predicting profit before a job gets to the shop is a critical phase of manufacturing. It begins by computing an accurate price for each job.

The Benefits of New Technology

An exciting change in computer-aided estimating is emerging. Some shops already have the capability to receive a request for quote (RFQ) electronically with an accompanying drawing. The estimating program can automatically recognize the part features and develop a suggested production method for the part. The estimator can review the information, accept or modify the process as needed, and send the customer a price for the part. All of this might take place in a matter of minutes. The customer can then accept the quote's cost - having reviewed the e-mailed information - and can reply with a confirming e-mail.

The shop, upon receiving the confirming e-mail, can forward the information from the estimating system to the shop management system and to the part programming system for processing on the shop floor. Realize as well that this opens up the estimating process to a whole new department that has always asked, "If I make this change, how will it affect the cost of manufacture?" Now design engineers using estimating/design tools as described above can make changes to the design and reprocess the part in the estimating/part recognition system, where they can see the instant impact of their design changes on part cost.

Those who have treated quotes lightly "because it is just an estimate" might lose the job before they begin their quoting.

Remember, when it comes to estimating, you don't care that it should take one minute - you need to know how long it will take. In an estimate simulation, the software attempts to include everything significant that occurs, as opposed to what should happen. The simulation process must factor into its calculations many issues: How many types of machines are used today by shops? How many different styles? How many different vendors? Rather than providing generic standards for turning or drilling, intelligent simulation will recognize perhaps 60 independent machine types for those processes.

What was said earlier bears repeating: The important fact to remember when buying estimating software is that each time you send a quote to a customer, you are betting your business on the accuracy of the times in the estimate.

Feature Recognition

The success that the solid modeling markets have seen - apart from the steep price decline in the past few years - is obvious: even someone in accounting can look at the image and recognize the part. They do not need an engineering background to be able to "see" the part. And from that solid, it is possible to make G-code, since you can recognize all of the features - the holes, flat surfaces, radii and so forth. As a result, most CAD/CAM systems now either integrate with a solids package or have their own internal solids drawing capability. Now, feature recognition is taking part design in a new direction. Solids packages enable feature recognition by other software - no more visualization, no more isometric views. Clusters of software packages that use solids will revolve around a technical database. This supports a consistent method of programming and manufacturing. In that technical database will be relationships of features to manufacturing methods. For example, a database entry for a 1/4-20 tapped hole, 3/4 deep 6061-T6 aluminum may specify that a .201 diameter drill be used and a .125 center drill spot that hole prior to drilling. When you machine a part out of 316 SS with a certain finish, the technical database can suggest that you machine using the specified speeds, feeds and tools that you have determined are appropriate. This makes everything consistent - this is the way that hole will be made, because this is the rule you set up. While this is the way it should be done today, on this machine, it doesn't mean that you can't change it. But it helps you to quickly and efficiently identify how a part should be made. Did you wonder why "material" was mentioned in the previous example? Why is it important that the material type be contained within a solid design? Because knowing it there can provide the part weight, center of gravity and possibly highlight other potential design or manufacturing challenges. That one piece of information significantly enhances a computer-aided estimating system's ability to create an estimate automatically. Solids are already at the next level today, where finite element analyses can help designers make parts faster, stronger and more efficiently. A solids program gives you the power of information.

Feature Recognition and Estimating

It won't be long until an estimator - knowing a part was designed in a solids program - will click on a pull-down menu and choose "recognize features." When the software recognizes a "shaft adapter," and the estimator selects the machining center to run the part, the software will suggest a center drill, a drill, a face milling operation, a contour endmilling operation and a burnishing operation. That's the set of work instructions based on the features it saw. The estimator will be able to drag and drop and rearrange any tool or process, based upon his knowledge of the facility. The point here is that this is a technology tool and will still require experience to know what can be done. Feature recognition in computer-aided estimating has already appeared in the marketplace - but so far, only for fabrication equipment. But it's not far behind for turning and milling. Consider a simple, flat part, linked from a solids package to estimating and then to a CAD/CAM system. The identified features from the solids technical database are matched with the estimating program's technical database. With this data, the estimating program recognizes the part and its material, and knows which tools to use. Perhaps there are several machines on which the job might run, so the estimator selects the machine. By automatically reviewing the machine parameters and the part, he may find that the initially selected machine is unavailable, or otherwise not optimal for the job - perhaps the first choice has insufficient tonnage, tolerances or size limitations. Therefore, the estimator could select another machine.