FINAL DEMONSTRATOR DOCUMENT for IDENTIFICATION SYSTEM FOR THERMAL DESORPTION SAMPLE TUBES (ID-DESORP) MARKES INTERNATIONAL LTD.

Transcription

1 FINAL DEMONSTRATOR DOCUMENT for IDENTIFICATION SYSTEM FOR THERMAL DESORPTION SAMPLE TUBES (ID-DESORP) MARKES INTERNATIONAL LTD. Associated TEC: UGCS Ltd. Page 1 of 12

2 1 Company Data Markes International is small UK company based in Llantrisant, South Wales. The company currently has 21 employees. Markes International was established in 1997 to serve thermal desorption users and associated scientific communities involved in the monitoring of volatile organic compounds. Thermal desorption is an extension to the analytical technique of gas chromatography and is commonly used in conjunction with it. In the process of thermal desorption, heat and a flow of inert gas are used to extract volatile and semi-volatile organics retained in a simple matrix or on a sorbent bed. The analytes desorp into the gas stream and are transferred into an analytical system in a small, concentrated volume of vapour. The company has an international client base which includes major industries, key regulatory agencies and the service laboratory sector with applications ranging from environmental health and safety to materials testing and product quality control. Amongst the applications areas of thermal desorption are: Industrial Hygiene: monitoring of exposure of workers to vapour phase organics. Environmental monitoring: ambient urban air, indoor air, emissions from building materials and domestic products, and tracer gas tests of building ventilation. Materials quality assurance and control: residual solvents, monomer and other materials. Organic composition: water-based paints, syrups, resins, medicinal ointments and others. Organic artefacts on silicon wafers and related materials. Markes International offers a comprehensive product portfolio including thermal desorption instruments, related sampling accessories, multi-tube conditioning rig and a wide range of consumables. The company designs and develops its products in-house with parts of the manufacturing contracted out. 2 Purpose of the project The purpose of this project was to develop a system for labelling and identifying industry standard sample tubes, which are used in the thermal desorption analysis process. Currently each sample tube is identified with a serial number which has to be physically stamped onto it, and with use this serial number becomes difficult to read. The increased trace ability and accountability will expand Markes International s sales, and give the company an advantage over the company s competition in the area of desorption analysis process equipment. The project has utilised passive radio frequency identification (RFID) technology, based on the use of standard components and microcontroller technology. Through the application of this RFID technology, the company has been able to respond to its customer demand for improved traceability of the test sample tubes and to establish an in-house knowledge of the technology for future applications in its analysis instruments. The main technical objectives of the project were; 1. The development of a RFiD read/write module which is to be embedded into the current product, to communicate with the encapsulated tags which will be mounted onto the sample tube. 2. The development of a standalone read/write unit which will allow the entering/updating of sample data in the field. Page 2 of 12

3 3 Existing Product 3.1 Product description The company developed its Ultra TD product, which is an advanced thermal desorper capable of handling 100 tubes located in 10 separate trays. This electronically controlled product is capable of analysing, industry standard sample tubes, and is compatible with other gas chromatography instruments, and is capable of handling volatile and semi-volatile organic compounds. The Ultra multi tube autosampler for the automatic desorption of up to 100 industry standard tubes, which are loaded 10 tubes in each of the 10 trays. Photo 1 Ultra TD auto sampler Page 3 of 12

4 The Ultra multi tube autosampler is controlled by software running on a PC, which allows the creation of sequences of desorption methods, which can be stored and recalled for future use. These sequences comprise of the time and date of each process, and deviations to the sequence are recorded, e.g. tube missing, gas leak failure, etc. 3.2 Technical description The functionality of the Ultra autosampler can be described by the following; Sample tubes are manually loaded into the 10 sample trays. All the sample data, particular to each tube, has to be manually entered into the sequence builder software. The Ultra, via an automated handling system, lifts each sample tube from its tray and locates it into an electrical heater assembly. The tube is heated to 300 C (typically) while an inert carrier gas is injected into the tube. The resultant desorped sample is then fed to a Gas Chromatograph for analysis. The system consists of 4 stepper motors, each with its own control board, and a heater with its associated control board. All boards are connected via a system bus, with overall system control being achieved by a central microcontroller which controls the sequencing and timing of the motors and heater unit. The user can modify these parameters via software running on a PC, which communicates with the central microcontroller. 3.3 Identified potential for enhancement By utilising RFiD technology, the improved traceability and reduction in user error can be achieved, by automating the sample data entry. As Markes International are the only company in this market to use the technology it will give them a considerable advantage over their competitors. An additional sales benefit is the sale of the tagged sample tubes, which is a new income stream to the company. 4 Product enhancement New product 4.1 Technical description The objective of the ID-DESORP project was to utilise radio frequency identification (RFID) technology to develop a system for labelling and identifying the sample tubes used in thermal desorption analysis. This will provide customers of the existing product, the improved traceablility of test samples which they require, and will establish in-house knowledge of the technology for future application within its analysis instruments. The system required the following features to fulfil the requirements of Markes International customers : Each sample tube has a unique number. Details of sampling parameters are stored with tube, these include; Identification number Date Period (start and stop times) Type (diffusion or pump) Volume (in litres) Location Operator Page 4 of 12

5 Tube trays can be filled in any order, as all sample details are stored locally with the tube. This removes a large source of error, as operators previously had to load trays and carefully record all sample details into a worksheet. Possibility to spilt a sample, and copy data to another tube. Run a scan of sample tubes, and store data in a worksheet. Provision of a small read/write unit to be controlled via a PC. This would be located near a sample site and used to input sample details prior to the tube being sent for analysis. System description. The analysis system consists of a sample tube tray handler, tube elevator and oven. Each tray contains 10 sample tubes; these are loaded manually and lifted by the unit to be loaded into the System Bus Power Base station IC Antenna RF Field Transponder Data Interface to Instrument Bus Microcontroller Read/write Module Figure 1 RF ID read/write module within instrument. oven. The system comprises of the existing analysis system, a reading unit (board), a transponder that is attached to the sample tube and a standalone transceiver unit. Photo 2 Embedded RFiD read/write module Page 5 of 12

6 The read/write module s antenna, which is to be mounted onto the gas inlet mechanism, ensures that the other tagged tubes are far enough away so not to interfere with communications. The base station IC is the Atmel U2270B IC that has a simple interface to the PIC 16F873 microcontroller, which controls transmission and reception. The module will interface to the instrument bus and will have an address set (during manufacture) to ensure that the correct commands are decoded by the required board, as the system has boards providing motor and heater control. The RFID read/write module acts as a dumb unit responding only to the commands issued by the master microcontroller. Sample tube transponder. Sample Tube End Cap Transponder Figure 2 Transponder location The transponder is mounted onto a sample tube, and read by the analysis system, via an antenna that is connected to the RFID read/write module. Data stored, which was programmed at the sample- gathering site, is sent by the transponder when the read/write module creates a RF field. The mounting of the read/write coil was to be such that it can reliably communicate with an individual tag, but not too close to adjacent tags to cause communication collisions. The electronic transponder tag is attached to the tube as shown in the illustrations above. When the tube is loaded on to a tray the tag is read before the tube is heated by the oven. A gas tight seal at both ends of the tube are implemented for pressurising the tube. There after the thermal desorption method is applied and temperatures in the oven will range from 200 C to 350 C. Standalone RFiD read/write unit. The features of the standalone unit are: Ability to read and write to transponder in the field. Units to be located near sample sites, connected to a PC or laptop. Data to be entered via a PC, application to be user friendly. Data can be updated, after sampling has completed. Embedded RFiD module to be utilised, with an interface board. Page 6 of 12

7 Power Base station IC Antenna RF Field Transponder PC interface board Data Interface to Instrument Bus Microcontroller Read/write Module RS232 to PC +9V DC Figure 3 Standalone RF ID read/write unit. The standalone unit utilises the RFiD read/write module which has been developed for the Ultra TD autosampler, and an interface board as been developed to convert RS232 serial data to the format of the Ultra s proprietary internal bus. The interface board also provides the power for the RFiD read/write module, via an external power supply. Technology choice The need for an effective labelling system for thermal desorption and gas chromatography analyses has been in demand by the users of these systems in a wide range of industries to achieve traceability of the analysis samples from the collection point to the analysis instrument. The use of barcode and numerical labels is not desirable for this type of applications because the information relating to the samples must held in a computer system or hardcopy form, with significant potential for making mistakes about the origin of the sample and the type of analysis required. Prior to the ID-DESORP project, Markes International did not have any expertise in electronic identification systems, including RFiD. This user experiment has enabled the company to acquire knowledge and expertise in deploying this technology in thermal desorption applications using best practice methodologies. Electronic methods of holding the data relating to the sample with the sample tube itself resolves this problem and provide the required traceability. The use of active data carriers are clearly undesirable because of the need for batteries. Therefore, passive electronic data carriers represented the best approach. The option chosen for implementing such a labelling system was RFiD technology, as the cost of standard read/write RFiD transponders are reasonably low because of their high volume production for other applications such as the automotive market. The standalone unit and the embedded read/write module embedded in the analysis instrument have the capability of communicating with the transponder without the need for contact. In this application, only very short distance communications is required (few millimetres) and therefore, the standard 125KHz FDX system has Page 7 of 12

8 been utilised in the system. Higher frequency RFID system are more expensive and have more complex reader circuits. The Atmel TK5551 transponder, which includes both the RFiD transponder chip and the antenna coil, is encapsulated in plastic and was attached to the sample tube by a custom made clip arrangement. A patent has been granted for the design of the clip and its application. The Atmel U2270B base station IC was used to communicate with the transponder device, providing both read and write capability. This device was controlled via a Microchip PIC microcontroller, sending and storing the data to and from the transponder device. The microcontroller also communicates to the central microcontroller, via the system bus, which issues read/write commands and receives or sends data to the RFiD board. The development process can be described as below; 1. Identification of the most suitable 125KHz encapsulated transponder, the Atmel TK Identification of suitable base station IC, the Atmel U2270B. 3. Prototype development of the RFiD read/write hardware, based upon the U2270B and a Microchip PIC16F873 ( a flash based 8 bit microcontroller). 4. Software design and development, which was written in C, and debugged using the Microchip In-circuit Debugger. The compiler used was the CCS PCM 'C' compiler, together with the Microchip IDE. 5. Interface board development, for PC to instrument bus connection. 6. Development of the PC based user data entry and reader software, which was developed using Borland Delphi Studio. 7. Printed Circuit Board development, of both RFiD and interface board. 8. Mechanical design and system integration. 9. System test and evaluation. The subcontractor chosen for the project was UGCS Ltd., as they had the relevant experience in RF design, embedded software design and PC application development. They also provided assistance in the recruitment of an Engineer for the project, who unfortunately left the company during the project. Through our collaboration with UGCS Ltd., the company have gained expertise in the area of RFiD and of the design issues relating to the use of the technology. Page 8 of 12

9 5 Time plan The original time plan and actual time for the project is shown below: Task 1: Technical Management Gantt Chart Project Months : Specification 3: Hardware design 4: Software development 5: Test and evaluation 6: Dissemination The actual time plan matches the original plan very closely, however there were a few changes: Specification an additional month was taken in this phase, where the decision to make a universal RFiD board, for both the embedded and standalone unit was defined. Hardware design the effort matched the planned time plan, but this phase started a month later than planned. Test and evaluation an additional month was required, as mechanical modifications to the existing product took longer than expected. Page 9 of 12

10 6 Budget The total budget predictions seem to have been reasonably accurate, and there was only a very minor deviations from the budget plan in terms of overall costs. Task Planned Effort (person days) Actual Effort (person days) Technical management System specification Hardware design Software design Test and evaluation Dissemination 8 5 total The expenditure under other cost categories were as follows: Cost Category Planned Actual (Euro) (Euro) Subcontractor Costs 21,900 20,706 Component Costs 4,000 0 Equipment Costs 2,500 0 Travel Costs 1, Total 29,800 21,026 The overall project development cost was approximately 55KEuro. Deviation from the original budget plan. Component costs - these costs were factored into the subcontractors costs. This included the build of the prototype and were part of the subcontractor negotiation process. Equipment costs the company did not purchase any development systems, existing systems were used. Travel costs the original budget included the airfare to travel to the final review within Europe. 7 Best Practice extraction 7.1 Problems, barriers and solutions The main barrier the company had to overcome was that there are very few suppliers of RFiD read/write ICs, which limited the choice of device selection. The requirement was for a device that operated at 125KHz, but had an interface to a microcontroller. There are few examples of this type of application of RFiD, applications tend to fall into the car alarm/immobiliser and logistics areas. The mechanical design challenge for the project was also a barrier, as there was the requirement to modify both the instrument and the sample tube. Through the evaluation phase different antenna Page 10 of 12

11 positions were tried, together with different tag positions. The resultant positions offer the best read/write performance together with the minimum modification to both instrument and sample tube. 7.2 Lessons learned and future impact The design of the read/write module was done is such a way that it can be used in both the instrument and the standalone unit. This required an additional interface board, which converts RS232 serial data to the proprietary bus format. This has enabled the company to maintain a small inventory of electronic components, and simplifies manufacture. Finally, a major lesson related to the selection of a reliable subcontract partner was learned when the recruited engineer left the company just prior to the completion of the development activity. Although the company will recruit another engineer, their role will now be to take over the maintenance of the current design and to bring the system to production. The present subcontract partner has provided assurances that this support will be delivered, and thereby ensuring that the company can introduce the new product to the market successfully. A major lesson for small companies is that the system knowledge required to support such a project is difficult to acquire, and back-up options to support the development must be considered. The future developments of the technology will concentrate, in the short term, on bringing this new product to market. This will involve the recruited engineer undertaking training, production planning and organisation and the management of system testing and commissioning. This will also involve development of a user guide, and training the company s other engineers. 8 Market situation, sales and market shares The company s market can be categorised into 2 main areas, viz. manual analysis systems and automated systems. The company s competitors in the manual systems area are entirely USA based companies, and these companies (Dynotherm, SIS, Teckmar etc.) are of a similar size and scale to that of Markes International. The company has entered the automated systems market with its Ultra TD product, and the major competitor in this market is Perkin Elmer (USA). Markes International has a 15% share of the worldwide market, and a 50% share of the UK market. The worldwide market has an estimated value in the region of 25MEuro. None of the company s competitors use RFID technology, and the adoption of RFID technology by Markes International will therefore provide a major competitive advantage for the company. The improved product from, offers the following advantages over competitors products: Traceability of tags, from the sample site through to analysis. Removal of the possibility of user error in data entry, all tags are read by instrument, and parameters set accordingly. Improved quality procedure due to increased tracking capability. There is also an increased revenue from the sale of the tagged tubes, which is in addition to the increased sales of the new instrument. There is also a possibility of supplying a retro-fit option to existing customers of the Ultra TD, further increasing the market for the product, and its associated tagged tubes. Page 11 of 12

12 Sales projections for both the existing and new product are shown below, values are based upon 2002/2003 figures. Sales Prediction 170% 150% Increase in sales 130% 110% 90% Exisiting Product New product 70% 50% 03/04 04/05 05/06 06/07 Year 9 Return on investment Figure 4 Sales predictions of existing and new product. The return on investment calculation is based on the cost of development and the cost of bringing the product to market. The development cost for the project is detailed previously at approximately 55KEuro. The estimated cost of bringing the system into full production is 20KEuro. This cost includes company engineer time, management time, staff training and continued sub-contractor support. The total costs to market will therefore be 75KEuro, the projected sales margins on the product is summed over 3 years and provides a Return On Investment (ROI) over this period of 4 years is 858%. Page 12 of 12