The Ancient Monuments Laboratory Bar Code Location Recording System

The AML bar coding system was developed in 1991 specifically to record the relocation of the bulk finds store from west London to Nottingham. This meant that the 2 main constrictions on development were that it had to be ready in time for the move (about a year after the project began) and, because of the lack of time, it had to fit in with the existing AML recording system. Since then major alterations have been made to take account of changes in location recording practice and changes to the database.

The Ancient Monuments Laboratory is a half way stage for material between excavation and deposition in a museum. Since virtually all the material dealt with is archaeological (apart from some decorative items from historic houses) most of it is small enough to bag and box making labelling fairly easy. The AML does not mark material. If this is done it is done by the excavators or the museum staff. Anything that does not fit in a box has a label tied to it.

Hardware (click on the highlighted items to see a picture of the piece of equipment)
TEC B-30 bar code printer (using thermal transfer printing option) [72Kb image]
Symbol 3800 Laser Data Terminals [84Kb image]
Symbol Charging/Communications cradle for above [63Kb image]
Symbol 2000 series bar code scanners [38Kb image]
Sequent Balance 8000 (old database server) [79Kb image]
Sun Sparc 10 (new database server)
PC

Software
Informix RDBMS (version 3)
Oracle RDBMS (version 7)
LBS label design software (supplied by TEC)
Bespoke program for Laser Data Terminals (written by Endata Ltd)
Microsoft C compiler

Printer media
110mm x 40mm polypropylene tag
50mm x 26mm paper labels

Previously all material which was sent to the AML received an accession number which was recorded on a central database along with the details of which site the material came from, context details etc. The current location for each accession number was also recorded. One accession number was given to each small find and each box of bulk finds. The accession number and other details were written on a card tag

When the decision was made to relocate the bulk finds store it decided that the quickest way to record the new locations would be to bar code the material and record the new locations with the scanners. All the material in the store would have to be checked (and much of it repacked) before it could be moved and it involved very little extra effort to print the accession numbers of the material onto the paper labels and stick these on the boxes while the checking was taking place.

The paper labels were chosen because they were a convenient size to hold a code 39 bar code and an ordinary version of a 6-8 digit accession number. Code 39 was chosen because it allowed all alpha-numeric characters plus some more of the ASCII charater set to be represented in bar code format. The labels needed to be of a high quality to go through the printer and had a synthetic adhesive on them. It would have been preferable to get acid free labels but it was no possible to find any which would go through the bar code printer. Also, the labels were going to be stuck onto unbuffered acid free boxes which therefore would not remain acid free themselves.

The printer uses thermal transfer printing which is of archival quality. A laser printer might have been an alternative if only labels were required but the TEC was needed to print the polypropylene tags. Later research by Austrailian museums has suggested that laser printers may not use sufficient heat when fusing the carbon particles to ensure a really permenent bond. (Australian Archives - Photocopying and Laser Printing Processes - Their Stability and Permanence")

Once all the material was packed at the old store it was transported to the new store at Nottingham. [Click on the image below for full 60Kb version]

Since the bulk finds store was bar coded all new material sent to the AML was given a bar coded as well. All material which arrived at the AML had the data entered on the database and then a polypropylene tag with the accession number in bar code form along with the rest of the data that was written on the card tags was printed from the database.

Once the bar code recording system was in use with the small finds the sheer numbers of objects in each box began to cause problems. Each time the movement of the material in a box had to be recorded it was necessary to open the box and scan the bar code of each object in the box. Previously there had been lists of contents taped to the boxes and these had to typed in by hand. Neither system was acceptable - the lists were often incorrect because material had been taken out or added to the box and the changes not recorded on the list and the bar code scanning was increasing the handling of delicate material. An upgrade of the database hardware and software provided a chance to redesign the system.

Now all material is put into a numbered box even if it is bulk material and there is only one accession number in the box. (To avoid the problem of having to relabel all the boxes currently stored any box which has not been assigned a number is given the number "UNKNOWN"). The new Oracle database records locations as either internal to the AML in which case the location must be a combination of room, shelf and box number already stored in a lookup table or a institution or person name stored in other tables. We are having a new program written for the LDTs by Peak Technologies (who have taken over Endata). This uses the data from the AML, institution and person lookup tables in the database as lookup tables on the scanner. The data has to be extracted from the database with SQL, transferred to my PC and then uploaded onto the scanners. The scanners then collect the new location information and the new data files are downloaded onto the PC and transferred back to the database via awk and SQL in a similar way to the old system.

Advantages

Extremely quick. There were c12,000 accessioned items in the bulk finds store. It took the equivalent of 1 person week to record all the new locations on the scanners and transfer this to the database. Without the bar coding system the new locations would have been recorded on paper and the details typed into the database. This would have taken something like 8.5 person weeks.

Accurate. The scanners either scan the bar code correctly or do not scan it at all. The system gets rid of all the typos that occur when numbers have to be written down and then typed in from the written list.

Less handling of fragile objects. The new system means that delicate objects do not have to be taken out of boxes to have their accession numbers checked each time they move around the AML. However all accession numbers have to be individually checked before the material leaves English Heritage.

Equipment. The LDTs have proved to be very robust. The type used has LCD display and integral keyboard (useful for entering data such as current room). They can be operated with one hand which is very convenient if the user has to climb up and down ladders when working in the stores. The LDTs are the type of scanner which use a red laser to scan bar codes from a distance. The other 2 types of scanner technology, wand and contact scanners, were rejected because they would have meant applying pressure to a bar code which was in a bag with object and therefore too the object as well. Also it would have meant having to get much closer to each box to scan it than is necessary at the moment.

The LDTs will also read surprisingly dirty bar codes and can also cope with bar codes in fairly dirty plastic bags as well. They can even read bar codes through thin polythene boxes.

Improved labelling. Using printed labels and tags has meant that the labelling of the material is tidier, easier to read and less likely to have mistakes in it than when it was handwritten. Also the labelling materials are better quality than were used before. The tags are especially are an improvement since they are inert unlike the card and far more robust. They are less likely to tear when tied to large objects through a hole punched in the tag and will actually survive the freeze-drying process complete with the thermal transfer printing although some slightly messy afterwards. (This also means that the conservators do not have to waste time printing out steel Dymo Tape labels for material that is going to be freeze-dried).

Streamlining of recording process. Since tags are now printed at the AML for the English Heritage Central Archaeology Service to use on site as well. Previously the finds staff on site had to write out card tags for each object found, enter the data on their database and then send a copy of the data to the AML along with the material. Since the CAS database allowed duplicate accession numbers and the AML one did not there used to be problems transfering data between the two databases. Now the CAS are provided with tags which have the site name and accession number in bar code and numerical form on them. All the finds staff have to do is to add the context, material and classification to the tag and enter the data on their database. Since there is only one tag for each accession number it gets rid of most of the duplicate recording of accession numbers.

Disadvantages

Incorrect labelling. Every so often people stick the wrong bar code on the wrong box. There is no way to prevent this and once it happens it is impossible to check for it until you actually find the box in question. There is nothing like discovering what the database said was a box of animal bone from Camber Castle is actually a charcoal sample from the Tower of London.

Label Design software. This was useful for learning what sort of labels the printer could do but could not do the one thing which was really essential - to take each record from a data file and create a unique label for it. This problem was solved be embedding the printer control codes in database report programs or simple C programs (eg for printing the CAS site tags).

Equipment. The 2000 series scanners have turned out to be of limited use. These are the type that plug into the keyboard of a PC. When the system was devised it was thought that it might be useful to have scanners on peoples' desks. However apart from making lists of objects in boxes they are not much use because they are not portable. Anything that it is possible to do with these scanners can be done with the LDTs with the right program.

The bar coding system is a vast improvement on the previous system. What the development of this system has shown is that using bar codes to record locations will not improve a badly designed system. The first version of the bar code system was designed specifically for the store relocation and worked there but would not work anywhere else in the AML because the recording system itself needed to be improved. The AML bar coding system has taken a long time to develop but now that it is working it would be impossible to run the AML object recording system without it.

Return To:
AML Home