All at Computing Laboratory, University of Kent at Canterbury, Canterbury, Kent, CT2 7NF UK Email: N.S.Ryan@ukc.ac.uk, J.Pascoe@ukc.ac.uk, D.R.Morse@ukc.ac.uk
Cite this as: Ryan, N., Pascoe, J. and Morse, D. 1997 FieldWorker Advanced 2.3.5 and FieldWorker Pro 0.91 GPS software, Internet Archaeology 3. https://doi.org/10.11141/ia.3.4
For further details, including price and how to order, see: http://www.fieldworker.com/
Until quite recently, field data recording directly into a hand-held computer has been a minority practice in archaeology. This is equally true of many other field-based disciplines and professions. Often there has been a very small number of enthusiastic individuals who have successfully used hand-held machines since they first became available. The majority have looked on, some making encouraging or even envious noises, but few have been tempted to follow the pioneers' lead. Most have tended to regard the available systems as too limited, too delicate or too expensive for their needs.
One area of field recording where hand-held computers have become established is automated data logging. Many archaeologists will have encountered direct data logging programs used with geophysical instruments, electronic theodolites or GPS survey equipment. Most are designed to work with, at best, a limited range of instruments, and many run only on a specialised hand-held computer, often one supplied with the instrument. When such software is available for more general purpose machines, it is often produced only for laptops running DOS or Windows, although the Psion hand-held range has established a small presence in this area.
Now that the desktop and laptop computer markets are well established, and possibly beginning to approach saturation, some manufacturers see hand-held devices as a major future growth area. The introduction of the low-cost PalmPilot by USRobotics and the new generation of machines designed around the Windows CE operating system are just two recent manifestations of this trend. Other manufacturers, such as Apple (now Newton Inc.) and Psion, have also responded with impressive developments of their existing ranges. Parallel improvements in communication technologies, particularly GSM mobile phones, have led to many of these devices being promoted as the latest form of executive toy; the personal web browser or email terminal thus replacing their earlier positioning as electronic organisers.
Most current hand-helds have simple text editors or word processors, some have rudimentary 'database' or spreadsheet programs, and there are often third-party products that provide simple form-based data entry programs. Together these represent an adequate, if simple, basis for recording both structured and unstructured data in the field. Modern hand-helds are far less limited than their personal organiser ancestors. Many are programmable and some are capable of processing tasks that would have taxed a laptop machine of a few years ago. Most are still too delicate to survive being dropped on hard surfaces and will need careful drying if immersed in water, although shock-resistant and partially waterproof rubber 'boots' are available for some models.
Costs have been reduced, but certainly not enough for most hand-helds to be regarded as disposable equipment. Nevertheless, many people now carry these machines and, as they become more widely used, we anticipate that cost will become less of an issue because they will be increasingly regarded as normal equipment. Just as we now regularly include desktop machines in our budgets so, in due course, will we also budget for hand-helds.
Despite these developments, we have one major reservation about the suitability of many currently available hand-helds for field use. Most are miniature versions of the conventional laptop 'clam shell' design with the keyboard as the primary input device. Whilst this is ideal for use when seated in a meeting or on a train, it is much less suitable for use when standing or walking in the open. Their shape makes these designs difficult to hold in one hand when not actively entering data, and the keyboard is hardly an ideal input mechanism for that essential component of field notes, drawings and sketches.
In our experience, pen-based systems are preferable for field input. These can accept hand-written text input but, most importantly, they allow sketches to be made without a change of input device. Several Windows CE machines now provide a pen as a substitute for a mouse or other pointing device and this can be used for drawing, but it is still necessary to revert to the keyboard for text input. Although early versions were widely criticised in the computer press, the handwriting recognition of current Newtons works particularly well, even if it does require some practice to get used to. The Graffiti system used on the PalmPilot is even easier to learn and use. It is, however, less flexible than the Newton input system and does not allow deferred recognition of hastily written text or the easy mixing of text and graphics.
FieldWorker, the subject of this review, is a data collection tool that combines general purpose recording forms and GPS data logging on the Newton MessagePad range of hand-held computers. It is produced by FieldWorker Products Ltd of Toronto, Canada. FieldWorker 1.0 was introduced in 1995. Since then, the product has been developed largely in response to feedback from users. In their manuals, FieldWorker Products emphasise their interest in hearing from and helping users. Our experience (albeit as reviewers), together with comments we have heard from others, indicates that the company is responsive to its users.
FieldWorker is produced in three forms, Basic, Advanced and Pro. Restricted demonstration versions of each of these are available from the FieldWorker web site. This site also contains a considerable amount of material abstracted from the very clear tutorial and other manuals supplied with the software. The Basic version, costing $149 (US), allows simple structured data to be recorded together with a GPS-derived location. We did not test this version but, from an examination of the FieldWorker Sofware Comparison chart, it would appear to be suitable for relatively simple tasks such as recording photographs or very brief observations during fieldwalking or other survey tasks. In May 1997 we tested the current release (2.3.5) of the Advanced version, and we were also able to examine a pre-release (0.91) of the Pro version. The Advanced version is priced at $599 (US) but, at the time of writing, the price and release date for the Pro version had not been fixed.
Following the supplier's recommendations, we ran the Advanced version on a MessagePad 130, the latest of the original series of Newtons, and the Pro version on the recently introduced MessagePad 2000. The 2000 is an altogether more powerful machine, up to ten times faster than the 130, thanks to the 160MHz 'StrongARM' processor, and with twice the display resolution (480x320 compared with the 320x240 of the earlier designs). Both programs will also run on the earlier MessagePad 120 provided that it is equipped with version 2.0 of the Newton operating system.
FieldWorker accepts serial input from a GPS receiver using the NMEA 0183 protocol. This protocol was originally designed as a means of sharing information between marine navigation instruments. Standard NMEA data 'sentences' defined for GPS receivers provide information such as time, position, heading, velocity, quality of position fix and the identifiers of satellites used in calculating the position. When the receiver is used with a differential correction input, information about the source and age of the correction data can also be provided.
NMEA output is the norm for most inexpensive hand-held GPS receivers and for those designed primarily for navigational use. However, some navigational systems and many survey quality receivers use proprietary communication protocols that are capable of carrying a much wider range of information than is permitted by the basic NMEA standard. Some of these, but by no means all, can output NMEA data, sometimes simultaneously on a second port. A serious limitation of the NMEA protocol is that the standard does not define any sentences for sending the raw satellite 'pseudo-range' measurements to the computer. Unless this information can be recorded, differential correction by postprocessing the data is not possible. All corrections must take place on-the-fly, and this requires an additional receiver and antenna to receive broadcast differential corrections. This equipment can add significantly to the cost and bulk of the necessary field apparatus.
Our tests were carried out using two GPS receivers: an off-the-shelf Garmin GPS45XL and a pocket-sized device constructed in-house and based on a Trimble Lassen-SK8 unit. The GPS45XL is an inexpensive hand-held unit widely available from suppliers of marine, camping and other 'outdoor' equipment. The Trimble receiver is a bare circuit board supplied to the OEM market for integration into a wide variety of products such as high quality vehicle navigation systems. We did not carry out any tests with high precision survey quality receivers.
We tested the two versions of FieldWorker in a variety of applications including a bird census on Mull, and a survey of earthworks on and around the University of Kent campus.
The FieldWorker data model is based on projects. Within each project, one or more screens are defined for data entry. Each screen is a data entry form in which fields can be all of the same type or of mixed types. In the Advanced version a screen consists of up to seven fields; a limitation imposed by the layout of fields on the MP120/130 display. The supported field types are Number only, Numeric slider, Text/Number, Notepad, Checkbox and Picklist. The Pro version allows an unlimited number of fields per screen and adds several other field types, including Date, Time, Temperature, Voice, Sketch, Long alphanumeric, Multi-choice picklist and Formula. The temperature and voice types rely on the internal sensor and microphone of the MP2000, the sketch type opens to a small sketch pad and the formula type allows values to be calculated from other fields or GPS data.
Normally, projects and screens would be developed on a desktop PC. This is a straightforward task that can be performed with a text editor or a spreadsheet program. FieldWorker has its own data import and export facilities, so there is no need to use the Newton Connection Utilities or other data transfer programs to move projects and data to or from the MessagePad. All that is needed on the desktop machine is a simple terminal emulator that can send and receive text files.
Project development can also be undertaken directly on the Newton, and a project control window allows the designer to grant or deny permission to the user to modify each part of the project. This is a useful facility where adherence to a particular data format is required but, equally, the ability to modify existing screens and to create new ones as the need arises is essential for research data collection.
Before starting to record project data, FieldWorker must be configured to suit the attached GPS receiver. The 'GPS Preferences' display covers a variety of options, including whether the GPS connects via the serial port or a PC (formerly PCMCIA) card, which NMEA sentences should be used, whether coordinates should be displayed in latitude and longitude or in UTM, and whether to use a single GPS measurement or to average location from several observations. The receiver accuracy can be entered and, when using Differential GPS, the antenna height above ground, the maximum age of differential signals and the maximum acceptable HDOP value can also be specified. HDOP is the Horizontal Dilution of Precision, a measure of the quality of the position calculated by the receiver.
The preferences display also offers a choice of geodetic datums. In use, the GPS receiver is set to output positions relative to the WGS-84 datum and FieldWorker will then transform these positions to the datum specified on the preferences display. The range of datums is small compared to that offered by many navigational GPS units, but covers many of the more widely used by national mapping agencies around the world.
Although the British Ordnance Survey datum is included, we were disappointed to see that the British and Irish national grids were not offered as coordinate systems. We discussed this issue with Cindy Park of FieldWorker Products who suggested that they would be happy to add this option if there were sufficient demand. We understand that at least one GPS supplier in the UK has also requested this enhancement. Interestingly, a possible solution appears to be scheduled for a future release of the Pro version. The FieldWorker Pro Features in Development page discusses a user-defined grid system. By entering the grid parameters into a new dialog, any Transverse Mercator grid system could be specified.
In use, data recording is based on 'stations'. Within each project, these are the locations at which information is recorded. Recording can start with a set of blank screens or with a duplicate of the current station, if only a few changes need to be made. Once a new station has been selected, the GPS location and a timestamp can be recorded at any time. These are not recorded automatically when the new station is selected, so the user can walk around a feature until a suitable point is found, or wait whilst the program calculates an average from several measurements.
Archaeological subjects are normally stationary and, in most cases, can be recorded at a leisurely pace. However, in some circumstances such as fieldwalking it can be necessary to record several observations at a single point and to do so quite rapidly. Although we did not test FieldWorker in this type of work, our experience of conducting a bird census with colleagues in the environmental sciences is relevant here.
This application involved standing at specific locations and recording the species, distance and behaviour of all birds seen during a predefined time interval. FieldWorker Advanced only allows one set of data to be recorded for each screen at any one station. This is not, in itself, a major problem because each observation can be treated as a new station. Indeed, it can be advantageous. For the bird census, where the observer stayed in one place for some time, the record of location and time with each 'station' was valuable as it provided a timestamp for each observation and offered the possibility of averaging the locations to give a better approximation to the true position. Unfortunately, neither of these benefits would apply to the fieldwalker who might need to make several observations of different artefact types during only a brief halt.
The relatively slow performance of the MessagePad 130 caused problems with the census. Switching between screens and, particularly, between stations was often too slow to be able to keep up with bird observations. Whilst speed was critical in this application, archaeologists should be aware that performance on this machine can be somewhat frustrating - the word 'treacle' was often used in our discussions. With the Pro version running on a MessagePad 2000, these problems did not arise. Switching between screens and stations was almost instantaneous, and we were able to make use of the multiple record feature of this version. Any screen can be defined as 'multiple'. Multiple screens can be used as many times as necessary at each station to record a set of similar observations. The suppliers describe this as a 'relational' capability because the recorded data is presented in tabular form, keyed on the station identifier.
Point locations can be taken from the latest GPS measurement or, when working at a more leisurely rate, by averaging a sequence of measurements. With both the Advanced and Pro versions it is also possible to record locations without actually visiting them provided that their range and bearing from the user's position can be measured. Additionally, the Pro version allows triangulation from a pair of GPS-derived points. To take full advantage of these features, additional equipment such as a compass and basic laser rangefinder would be needed.
Whilst the Advanced version will only record single point locations for each station, it is possible to record a separate trail of points whilst walking along a linear feature or around an area. There are two ways of achieving this, but neither is entirely satisfactory. The first is to use the automatic trail facility to create a new station at regular time intervals or after a preset distance has been covered. This, of course, leads to a large number of almost empty station records because the observations would normally be recorded at one station and only the locations of the others would be of interest. Using this method it would probably be wise to avoid adding further observations at intermediate stations unless absolutely necessary as this would complicate the task of separating geometry and attributes after the data has been transferred to the desktop.
The second method is to record the NMEA output of the GPS receiver whilst walking the feature. Again this is not a very satisfactory method as the coordinates will have to be extracted from the NMEA sentences and transformations applied to convert them to the appropriate grid and datum. Simple mapping can be achieved by either of these methods but in both cases some programming, or considerable manual editing, effort will be needed to extract usable data after returning from the field.
The Pro version provides much improved support for recording linear and polygonal features. Instead of a single button to record the point location of each station, three buttons are provided to give a choice of point, line or polygon. In this way, the geometry of the feature is directly associated with a single set of station attribute data. This is an altogether more satisfactory approach that recognises the need of many users to collect geometric data for later input into a GIS or other mapping program. Compared with our experience of using the Advanced version, we found the task of mapping linear earthworks, enclosures and quarry pits to be quite straightforward.
Both Advanced and Pro versions include map and 'navigator' displays. The map shows the location of each recorded station in the current project, and a 'legend' button allows any data field to be selected as the station label. Tapping on a station symbol takes the user to the data entry screen to view the records for the station. The trail button toggles recording and display of the route taken. The grid size can be set to any desired value in units of miles, feet, kilometres or metres, and the usual zoom and scroll buttons are provided. North need not be at the top of the map: the small button with a north arrow symbol leads to an orientation dialog with which the map can be aligned on one of eight primary compass directions, or to an arbitrary bearing.
The 'navigator' display is used to show the range and direction to any selected station. The direction indicator relies on the NMEA VTG sentence, a message containing heading and speed data. A GPS receiver can only determine the user's heading when moving so this is not a complete substitute for a magnetic compass.
Although FieldWorker could be used as a stand-alone reference system, most users will want to transfer the data collected in the field to a desktop machine. Data can be exported to the desktop in a variety of formats. The basic FieldWorker format is similar to that used for preparing projects and it is a relatively easy task to extract any required information from it using a text editor or, more effectively, by using a simple program written in a language such as Awk or Perl. Fortunately, many users will not need to take this route as data can also be exported in formats compatible with several widely used GIS and mapping packages including MapInfo, ArcView, Fugawi and StreetsOnADisk. Location data can also be exported as UTM coordinates or in the format used by the Trimble ScoutMaster GPS receiver.
In some ways we were disappointed by the Advanced version, but rather than any particular deficiency in the software, this was largely due to the nature of the tasks we set it, the relatively slow performance of the Newton MessagePad 130 compared with the 2000, and our exposure to the Pro version. Despite its more capable map display and other features, several of the projects we undertook with the Advanced version were simple and could probably have been achieved adequately with the much cheaper Basic version. Others were far better suited to the Pro version with its ability to record details of linear and polygonal features and multiple data records at a single station.
The lack of direct support for British and Irish Ordnance Survey grid references will be seen by many in these countries as a serious limitation in current versions. However, as mentioned above, a solution appears to be in hand for future release, at least in the Pro version. This will not solve the problems of all national mapping grids, only those based on Transverse Mercator projections.
Most Newton software is remarkably inexpensive with many packages costing much less than £100 or $150. Judged in this light, FieldWorker Advanced seems expensive, but it is a specialised product with a relatively small market. Those who have used a simple hand-held navigational GPS such as the Garmin unit used in our tests might be tempted to ask why the program includes so many features that these receivers offer at a far lower price. The answer is obviously that FieldWorker is intended for use with a much wider range of receivers, many of which do not include navigational features such as waypoints and track recording.
Those on a very limited budget could achieve similar results with a hand-held GPS and a simple forms package, or even the built-in notepad, on a Newton or other computer. Admittedly, locations would have to be entered by hand from the GPS display, and extracting useful data from both machines would be more time-consuming and error-prone. The main problem with this approach would be the difficulty of carrying and referring to two hand-held devices. From our experience, earthworks might most easily be located by falling over or into them. With a purpose-written program like FieldWorker, the GPS can be kept in a pocket or pack, perhaps with a small external antenna attached to a hat, pack frame or ranging pole, and the user would only need to concentrate on the computer.
Our mobile computing research group already owns several Newtons and GPS units, so if we were looking for a commercial package to support our field activities, we might well consider the cost of FieldWorker to be quite reasonable. However, many field archaeologists would need to purchase a complete package consisting of a Newton MessagePad (either the 130 at about £400/$600 or the 2000 at £600/$1000), a GPS receiver (£200/$300 for a simple hand-held unit, rising to £10,000/$15,000 or more for survey quality equipment). With the simpler GPS units a differential beacon receiver might also be needed together, in many areas, with a licence to receive differential broadcasts from a commercial source (typical costs begin around £600/$1000). As only a part of a total package, the cost of FieldWorker becomes less significant, but whether many field archaeologists would see the total as worthwhile expenditure is a quite different question.
On balance, we find it difficult to recommend the Advanced version for use in archaeological field survey because it falls short of typical requirements in several important areas. For most users, the price could only be justified if the product met their criteria much more closely. The Pro version will almost certainly be more expensive, but it is definitely worth a closer look and may provide a valuable tool for many archaeological field survey projects. For those in need of a system to record simple point-based observations, such as a photographic record book, the Basic version will almost certainly suffice.
Internet Archaeology is an open access journal based in the Department of Archaeology, University of York. Except where otherwise noted, content from this work may be used under the terms of the Creative Commons Attribution 3.0 (CC BY) Unported licence, which permits unrestricted use, distribution, and reproduction in any medium, provided that attribution to the author(s), the title of the work, the Internet Archaeology journal and the relevant URL/DOI are given.
Terms and Conditions | Legal Statements | Privacy Policy | Cookies Policy | Citing Internet Archaeology
Internet Archaeology content is preserved for the long term with the Archaeology Data Service. Help sustain and support open access publication by donating to our Open Access Archaeology Fund.