averages for the time period. During the next time period, the system would collect
another data set and that would go in the next record of the file. This data transfer
would continue until the storage resources of the minicomputer were full and then
the process would start over with the first record of the file being written over the
data the computer had previously stored in that record. As has been mentioned
in earlier chapters, this type of data collection approach is called the circular file
approach, because the records are overwritten in a somewhat circular manner.
With this type of data collection system, the minicomputer’s bulk storage would
contain a snapshot of process data for the period of time the size of the storage
device would allow. It was not unusual to see systems set up like this with thousands
of process data values in each record and with enough records to cover the last week.
Since the communications were typically only from the DCS to the minicomputer,
and large amounts of process data were sucked across the computer gateway and
into the minicomputer, this approach was known as the vacuum cleaner approach
(Figure 17-5). The minicomputer was the vacuum cleaner sucking all the data it could
out of the DCS. Once again, specialty talent implemented these systems, and systems
integrators and the company’s DCS supplier began offering such services.
Unfortunately, a survey conducted around the time of these systems found most
of the data that had been sucked up into the minicomputer was not even touched by
the business system before it was overwritten. The reason was to the business systems
professionals, the data in the minicomputer was nothing more than useless raw data.
The automation professionals provided as much raw data as they could without much
context associated with the data, and the business systems professionals never really
understood the data nor did they know what to do with it. This initial attempt to
integrate the business and automation domains is important because it points out the
naïveté of the automation teams and the information technology teams about what
actually took place in the other’s domain. This naïveté is still very much in place today.