6. Digital Developments Applied and Assessed during the DigIT Excavation

A specific objective of the excavations undertaken as part of the DigIT project was the testing and assessment of digital technologies applied in the field, building on the methods and approaches applied in the various pieces of fieldwork undertaken at Heslerton since the late 1970s. New technologies, both hardware and software, offered the opportunity to either replace or enhance many of the approaches previously used in Heslerton.

6.1 PALM-based recording

When the DigIT project was in progress the PALM-based system was selected as being cheaper, slightly more durable (with fewer contacts that might allow sand or moisture to damage the instruments) and offered a far better software base that would interface directly with the Access database in daily use. At the time of writing the situation has changed a little, the Windows Mobile platform has greatly improved and the software base has likewise matured. The cost of PDAs, whether based on the PALM or Microsoft Windows Mobile operating systems, have continued to reduce since the DigIT excavation was completed, such that there can be little objection to their widespread use for primary recording on excavations.

Operation of PDA devices is quite simple, using a stylus on the small screen for data entry either through a cursive script – Graffiti in the case of the PALM, or using a keyboard to tap in letters which are then displayed in the text entry area, or through the use of pull-down lists of key terms. In the case of the LRC Context and Object records, the structure and keyword pull-down list fields were easy to import from our main Access database using The ThinkDB software. Setting up the data entry screen using a series of tabbed data entry pages took only a few hours to complete, and these combine both pull-down list fields covering aspects such as shape, planform, context type and free text entries. The use of pull-down list fields not only makes data entry fast but encourages consistency of data entry, ease and reliability of analysis and database searches. The list fields, which have a limit of 75 terms per field, were compiled from the terms in the thousands of extant records.

Even those less familiar with computers quickly became familiar with the use of the handheld PDAs. To avoid confusion and make daily management easier, only two devices were issued for each trench. The team for each trench was small and two devices could easily cover what was required. The quality of the record substantially reflects the archaeological experience of the recorder and it was felt that those who were recording for the first time should do so with the trench supervisor as part of a training exercise. The approach to excavation in Heslerton has always been heavily reliant on the area supervisors and site assistants (most of whom joined the excavation as volunteers and returned in later seasons to act as site assistants and supervisors) to work with the individual excavators and carry out the bulk of the recording. This approach has benefited the many hundreds of initially inexperienced volunteers who have worked on the excavations in Heslerton. This is an effective approach given the seasonal nature of the fieldwork, lack of long-term staff and the many novel techniques employed that were not generally applied elsewhere. In one excavation season during the 1980s, with many inexperienced excavators, an attempt to get everyone to do all their own recording rapidly descended into chaos. Since the context record is the fundamental cornerstone of the whole archive, it is essential that these are recorded either by or with someone of sufficient excavation and recording experience: the quality of information returned from an inexperienced volunteer through a 'questions and answers' session with the supervisor is far better than that from someone simply left with a form or PDA and asked to record what is there.

At the end of each day the records were backed up on a laptop in the site hut, where the data were imported into the main excavation database. This could then be checked on a daily basis to identify and track down information that for some reason had been omitted from the record. Regardless of whether records are gathered using pro-forma sheets or PDAs, daily record checking is essential to ensure that vital information is not omitted. Prior to the use of PDAs all software used for data collection was programmed within the LRC; this made it possible to employ simple verification techniques to send records deemed unfinished back to the handheld. It was effectively impossible to remove an incomplete record as these survived in memory even after the batteries were removed. This was not possible with the PDAs, meaning that more time was needed for the data verification process. The data verification generated error lists that could be handed back to the supervisors who, thanks to the radically improved storage capacity of the PDA, could keep the full record on the PDA. The LRC database structures are documented in Appendix 3.

In addition to using the PDAs for recording the context records, colour models were tested using a plug-in Eye-Module camera and with a digital drawing device, the Seiko Smartpad. The images captured with the Eye-Module were of very low resolution and except in bright sunshine were of poor quality (poorer than the Polaroid shots that we have used to enhance the record in the past). The Seiko Smartpad used the combination of a sonic pen and a small A5 format digitising surface to allow vector drawings to be directly entered into the PDA, which was mounted on the pad with Velcro and communicated with the digitiser using its infra-red port. The pen was fitted with a ballpoint pen tip and could be used to draw conventionally on draughting film, the drawing being digitised at the same time. This device showed real promise but was let down by the fact that the data could only be exported from the PDA as a bitmap image, rather than in scaleable vector formats suitable for incorporation in the GIS, alongside the difficulty of doing field drawings using a ballpoint pen, the lack of an erase function and the small sheet size. Since DigIT was completed the same technology has been incorporated into a clip that can be mounted on an A4 clip board and the vector data format has been released; we have not had the opportunity to try this out and a visit to the Seiko website reveals that this is a discontinued product.

In the case of the DigIT excavation the finds were all recorded at the LRC offices; although the PDA version of the finds record works well, it is easier to enter the records quickly using the Access database in spreadsheet data entry format.

Although the Handspring PDAs were durable, sticky-backed plastic screen protectors had to be applied to the glass screens to protect them from scratches (a potentially serious problem in the sandy environment of the excavation). The device could still be used in very poor and wet conditions when contained in an Aquapack case, a completely sealed protective case with a clear cover, although it was a little more difficult to write using Graffiti through the cover.

Improvements in local radio communications standards for mobile computing, particularly WiFi, Bluetooth and mobile telephone data communications, mean that alternative configurations providing direct shared access to a central database, running on a laptop computer or even an off-site server are now quite possible. The DigIT experiment was slightly ahead of the technology in this respect as, at the time, while it was possible to develop an online context record using Active Server Pages (ASP), the costs of interfacing the data collection in the field with the ASP-driven dataset was prohibitive. More critically, while it is technically possible to operate with an active online database while excavation is in progress, why would one wish to do this? Beyond a technical demonstration, this is highly questionable during excavation although we can certainly see the benefits of such an online record at the post-excavation stage, when many specialists, who may be based elsewhere in the country, could be given direct access to the record during the analytical phase. Although we see direct digital data collection in the field as a basic objective or even requirement, the importance of that data cannot be realised without the plans, photographs and other retained materials that make up the excavation archive.

As with any recording system, the application of the PDA instead of pro-forma sheets or site notebooks for primary recording does require careful management. The ThinkDB software used in the field has highly sophisticated functions for synchronising data from individual PDAs and the main access database. It was quickly realised that this synchronisation had to be applied with great care; in our first set-up, using the default settings, the system would import data from all the PDAs and update each with the synchronised data. However, it was immediately realised that this could allow new data in one device to be updated with old data held in another. It was, however, a simple task to restrict the updates from the main database so that active data was not overwritten. In order to be absolutely sure that there was no risk of false reverse synchronisation, exporting the synchronised data back to the handheld was not employed, and the PDAs were assigned to a particular Site, Area and block of context numbers, meaning that multiple PDAs could be used in the same area without any risk of data being lost during the update process. The sort of basic records management that was required was in many ways no different to that employed for many years on excavations recorded using pro-forma sheets, where blocks of context numbers were released per individual so that there was no risk of a number being used twice for two different contexts.

To the best of our knowledge, Nick Ryan (Department of Computing, University of Kent), was the only other archaeologist in Britain using PDAs in the field at the time of the DigIT project (Morse et al. 1998; Ryan et al. 1999a; 1999b). It remains remarkable that this basic technology has not been applied on a far wider basis. Rather than labour the point, it can simply be stated that the case for the use of the PDA in the field is proven; they are cheap, relatively durable, easier and more flexible than any handheld devices used in Heslerton during the previous 16 years.

PREVIOUS NEXT CONTENTS HOME