Extracting Machine Performance Information From Site-Specific Grain Yield Data To Enhance Crop Production Management Practices July 2001

R. Mack Strickland, Daniel R. Ess, Samuel D. Parsons, and Matthew T. Crisler

The estimated 30,000 yield monitoring systems on combines in the U.S. generate staggering quantities of data. The most obvious use of these data is to investigate yield variability in the field. Yield variability can provide the rationale for implementing site-specific management in crop production. The same data sets that provide yield information can also be “mined” for other information that can have significant management implications. That additional information includes travel speed, effective cutting width, and the breakdown between productive and non-productive time in the field. Special techniques have been developed by a team of Purdue University Agricultural Systems Management researchers to perform analyses of yield data files to provide improved machine systems management information.

In the past, field studies of machine performance were accomplished using devices such as stopwatches, clipboards, and record forms. These methods were time consuming and required a person or a team in the field when the machine was operating. Through the use of geo- and time-referenced yield data collected with a crop yield monitor, actual combine performance can be investigated without a large investment of time. And the investigations can be conducted at a time far more convenient than during the rush of a harvest season.

Effective management of machine systems becomes increasingly important as farm size continues to grow in the U.S. Combines of increasing capacity require improved machine system management in order to operate at optimal productivity levels. It is critical that the “supporting cast” of transport and support vehicles, and receiving, handling, and conditioning equipment be sized and situated to make the fullest use of combine capacity. The information embedded in typical crop yield files can help to identify factors that limit productivity in the field.

The Purdue team developed methods and algorithms to automate the extraction of detailed combine performance information from typical crop yield data files (Fig. 1). The extracted, detailed performance information can be used to aid crop producers in selecting properly sized equipment. It also is useful for crop producers and researchers to make combine performance comparisons among various fields and operators, combine makes and models, or alternative cultural practices.

* Monitor ID, Field Name, and Crop do not change as the crop yield data is recorded.

Figure 1. Portion of an Ag Leaderâ advanced format (.txt) yield data file.

The determination of effective working width is an important step in calculating actual (effective) machine capacity. The effective working width of a combine header is affected by the amount of overlap into previously harvested areas. Overlap can vary for different combine operators and field conditions and is difficult to determine visually. Effective combine header width for crops that are harvested in rows is the product of the row spacing and the number of rows being harvested. While the operator is required to enter a “nominal header width”, the actual effective header width can be determined using yield data files by calculating the distance between parallel passes across the field.

By selecting two approximately straight and approximately parallel passes in the yield data file (Figure 2), calculating the distance between the two selected passes, and dividing the distance by the number of passes between the two selected passes plus one, the approximate effective header width of a combine can be calculated.

Figure 2. Illustration of two approximately straight and parallel combine passes that could be used to determine effective header width.

Combine performance information such as field efficiency(the ratio of actual combine field capacity to the theoretical maximum combine field capacity) can vary greatly among combines, operators, and field situations. Self-propelled combine field efficiency typically ranges from 65-80%. Field efficiency is reduced when time spent doing non-harvest activities such as turning at end rows, idle travel, unloading grain, unplugging the combine, adjusting the combine, lubricating and refueling the combine, and waiting for other machines is increased. Whenever one of these non-harvest activities occurs in a crop yield data file it can either be identified by a skip in the time sequence (Figure 1) or when the travel distance of the combine is zero. A time skip in the yield data file occurs any time the combine header is raised for 5 or more seconds. The duration of the non-harvest activity is simply the difference in recorded time values (in seconds) between the beginning and end of the skip.

By summing the individual time skips and the time when combine travel distance is zero, the total amount of time spent doing non-harvest activities was calculated. By subtracting this value from the total field time (difference between the first and last time value in the yield data file) plus the time required to unload grain after the field is completed, the actual harvest time can be determined. By dividing the actual harvest time by the total field time, the overall combine field efficiency can be calculated.

While overall field efficiency is important, determining the efficiency losses resulting from different types of non-harvest activities can also be informative. For example, knowing the amount of time that is spent unloading the combine can be useful for basing management decisions about the type of unloading method or transport system that are needed.

Non-harvest activities that result in field efficiency losses can typically be separated into three major categories: turning at the end of a pass, unloading grain (if a stop-unloading harvest method is being used), and other non-harvest activities. Figure 3 illustrates these three major types of events and shows the time skips associated with them. Methods were developed to calculate the losses in efficiency resulting from these three major activities.

Figure 3. Example of time skip and pass information available in a yield data file.

Factors that affect combine performance or field capacity, such as combine travel speed, effective width, and field capacity can be extracted from site-specific, geo- and time-referenced crop yield data files. Information about losses in field efficiency can be extracted in addition to insights into efficiency losses resulting from turning, unloading grain and other non-harvest activities. These techniques can be used to expand a producer’s machine management information base with far less labor expenditure than was necessary with the use of traditional time-and-motion study techniques.