5. The Landscape Research Centre Field Recording System

5.1 The historical context

Archaeological excavations undertaken by the LRC, formerly under the banner of the Heslerton Parish Project, have been at the forefront of applied field computing since 1984, when handheld computers were adopted rather than paper records for recording context and finds data in the field (Powlesland 1986; 1991; 1997; 1998; Powlesland et al. 1986; 1998).

A computerised data structure for handling field records from excavations in Heslerton was developed in association with the North Yorkshire County Council (NYCC) during 1980-81. This first system was fully integrated into the County Sites and Monuments Record and provided a record structure that incorporated data from individual excavated finds to entire landscapes in a seamless structure. The one significant weakness of the system was the requirement that pro-forma recording sheets, filled-in in the field, had to be transcribed onto the computerised recording system maintained on the County Council mainframe, using a remote terminal installed at the excavation headquarters. The terminal was a simple text-based terminal; there were no local printing facilities and the line to County Hall, 50 miles away, frequently failed, requiring data to be re-entered sometimes as many as three times. We rapidly became aware that while it was perfectly possible to request printouts and distribution maps from County Hall, the turn-around time on getting these back to site was several days, and our printing and plotting requirements were far greater than had been anticipated by the local authority computing team.

The demands imposed by the rescue excavation environment, with limited funds and time available in the field, meant that records could often only be entered after the fieldwork was complete, by which time the site was in the process of disappearing as it was quarried away, making any subsequent record-checking against evidence in the ground impossible.

It was quickly realised that a better alternative would be to find a method of performing data entry on site, using a local PC for day-to-day data management, with the archives being transferred to, and maintained by, the County Council on its mainframe computer. This approach satisfied the local requirements and, in theory, should have guaranteed the permanent maintenance of the data archive within the local authority data management system, with inbuilt migration to new systems over time. This approach was adopted for a trial dataset relating to excavations at Cook's Quarry (Powlesland 1987). However, the local authority later deleted the archives relating to this trial when the Sites and Monuments Record was re-formatted into a somewhat reduced, and in our opinion poorer, record structure. Fortunately, the paper records survived as the trial data had been entered from field recording sheets.

By 1984 handheld computers, programmable using a form of BASIC, had become available. The Sharp PC 1500, launched in 1982, was durable enough for use in the field and easily programmed. These were used in Heslerton from 1984 until 1986 for all Context and Object recording in the field (the PC1500 continued in use for on-site finds recording until 1995, when recording was transferred to a laptop). This device, with a one-line 26 character display and only 8k of memory, required the extensive use of numerically coded data in order to maximise the use of the limited memory for data storage. Later models, including the PC 1600 (released in 1986), which had the same form factor but included a four-line display and up to 64k of memory on an expansion card, replaced the earlier model, providing greater capacity without the need for any re-programming. These remarkable devices supported a small pen plotter/printer, were reliable and durable, and formed the backbone of our recording system for many years. During 1987/8 some data, particularly survey data, were logged using the original PSION handheld computer, but in its first version it was found to be susceptible to data loss if knocked, causing poor battery contact, and the use of the PSION was abandoned. By 1989, when MSDOS was becoming the standard for personal computing we experimented with the Atari Portfolio, a PC-compatible device with a fold-down 8-line 40 character display; although we used these for one season they were not as durable as the Sharps and were replaced with the Hewlett Packard HP95LX, which ran a home-compiled version of dBase written in BASIC. Since 2001 the field recording system has been based on the use of Handspring PDAs running the PALM operating system and commercially available software, as discussed below.

Although some of the hardware was purchased, we were fortunate to receive much of the handheld equipment through charitable donation or sponsorship arrangements. We remain indebted to the various manufacturers: Sharp UK, Atari, Hewlett Packard UK and most recently to the Handspring foundation, for their support and assistance over more than 16 years. During that time hundreds of thousands of records have been put through the various handheld devices with minimal data loss.

Despite the limited storage capacity of the various handheld devices used in the field in the past, the data structures established in the early 1980s have continued to form the core of our recording system. The use of numeric codes for data entry was dropped during the 1990s, by which time the storage capacities of the devices had expanded and, once we were able to use MSDOS compatible devices, the records no longer had to be maintained in a home-built binary database and all data were gathered and maintained in dBase III format files, managed on the PC using FoxPro software.

PREVIOUS NEXT CONTENTS HOME