**4. RFID implementation for forest industry**

The past RFID trials have focused on using available commercial RFID transponders to mark logs or other wood items. The results of these trials have varied, but in general the transponders intended for other applications have not been optimal for the needs of the for‐ est industry. Therefore, a custom made RFID solution was considered advantageous and was developed in an EU FP6 funded project called Indisputable Key [10]. The following Sec‐ tions describe the passive UHF RFID solution developed for the supply chain of the forest industry.

#### **4.1. RFID transponder for log marking in sawn timber supply chain**

The basis of the traceability utilising RFID is the transponder used to mark the wood items. The requirements for the transponder to be used in log marking in the Nordic sawn timber wood supply chain can be summarised as follows:

**•** High readability

**3.2. Challenges of UHF RFID technology in wood traceability**

308 Radio Frequency Identification from System to Applications

system performance.

mum allowed radiated power.

er survival rate considerable.

Economically viable utilisation of the traceability requires that the wood items can be auto‐ matically identified. The item marking should not reduce their value or limit their use as a high quality raw material. The identification of wood items, the marking and reading, should be done without reducing the production efficiency e.g. by slowing it down and the costs related to the traceability should be reasonable to allow the benefits of the traceability to be utilised. The most significant part of the RFID system is the transponder as they are the most numerous component in the system and their performance is the basis of the overall

Wood is a natural material with varying properties between the trees, logs and boards – and within them. The density of the wood, the grain orientation and the moisture content vary and thus the electromagnetic properties (the complex permittivity) also vary. The varying moisture content has the greatest effect on the permittivity and loss in the wood. The per‐ mittivity variation can lead to transponder antenna detuning and the high loss due to the high moisture content attenuates the radio signal. These effects have to be taken into ac‐ count in the design of the UHF RFID transponder to guarantee a sufficient reading range in all conditions in the wood supply chain. UHF label tags are therefore not suitable for mark‐ ing fresh wood with high moisture content. In practice, the reading of the tags at saw mills has to be done at distances up to 1 m. The reading range depends mostly on the transponder as the reader operation is governed by the radio regulations defining for example the maxi‐

Reliable identification of the wood items requires that the tags have a high survival rate in the wood processing steps as transponders that have been destroyed or have been detached from the logs or boards cannot be read. In the RFID trials it has been frequently found out that tags glued, stapled or otherwise attached on the logs may be lost in the wood process‐ ing – especially during transportation, on conveyors and in debarking. In [9] it is reported that some 75 % of tags attached to the front-end of the log were lost in debarking at the saw mill. In trials carried out by the authors with tags attached onto the surface of the log ends typically up to a few per cent of the transponders were lost in each processing step which results in a significant loss of tags over the supply chain. Therefore, in order to ensure the transponder survival through the whole supply chain the tags has to be inserted inside the wood. Inside the wood the tag is protected from impacts which will improve the transpond‐

The transponder has to be attached on or preferably inserted into the wood by an applicator tool or machine and the tag has to be suitable for reliable and quick application. The applica‐ tion of the transponder should not reduce the production efficiency i.e. the application should not introduce significant delays. The application has to be done automatically where the wood processing is automatic and manual application is possible only if the wood is handled manually, e.g. felled with a chain saw or a reasonably small number of logs are

marked. The transponder has to withstand the application to be readable.


These requirements are discussed in Section 3. The required compatibility of the material used with the pulp and paper making processes is perhaps the most constricting require‐ ment for the transponder. Typically a UHF RFID transponder consists of a thin plastic inlay with a metal foil for the antenna to which the microchip is connected and of a hard plastic casing. As common plastics are not accepted in the wood used for pulping, alternative mate‐ rials were considered. Biopolymers offer an interesting alternative to conventional plastics.

In addition to the chemical compatibility with the paper making processes, the transponder material has to be suitable for insertion into the wood to ensure tag survival in the logs in the wood processing steps in the supply chain. The material has to be mechanically durable; sufficiently hard but not brittle and it may not absorb water. The transponders have to sur‐ vive several months in the logs. The material should have suitable electromagnetic proper‐ ties at UHF frequencies – ideally low loss and stable properties. In addition to the suitable chemical, mechanical and electrical properties the material has to be applicable for mass production of the transponders using common plastic fabrication techniques e.g. injection moulding. A suitable bio-composite material meeting these requirements is ARBOFORM® by Tecnaro GmbH [11] and it was selected as the transponder casing material. The ARBO‐ FORM® -material consists of lignin, natural fibres and processing aids. To facilitate the mass production, conventional plastic inlay with aluminium as the antenna pattern material was selected, as the amount of plastic in the inlays ending up into the pulping from the saw mill is negligible. Currently, paper inlays are also available for a non-plastic alternative. The transponder is EPC Class 1 Generation 2 compatible.

simulations, Ansoft HFSS was used. The transponder readability is best when the tag is in the end of the log, as this part is usually exposed in the piles and on conveyor. If the trans‐ ponder is in the side of the log it may be left under the log or covered by other logs and reading would have to be done through considerable thickness of wood and with the possi‐ bility of the tag being pressed against a metal surface. Figure 7 shows the simulator model of the transponder inside the log and the basic layout of the planar dipole antenna inlay.

the log and the basic layout of the planar dipole antenna inlay.

Figure 7. Simulation model of the transponder inside a log and the planar dipole antenna layout.

with the measured reading range using TagFormance™ measurement device.

The planar dipole antenna was optimised for operation inside wet wood with tolerance for varying permittivity caused by varying moisture content in the wood. The electrical proper‐ ties of wood were measured at UHF and the relative permittivity of the spruce was found to be of the order of 2.3 when fresh and 1.8 after kiln drying. Correspondingly, the loss tangent was 0.08 and 0.03 for fresh and dry spruce. When soaking wet, the relative permittivity of the wood may be even in the magnitude of 10. The final antenna design has the dimensions of 74 mm x 5 mm. Figure 8 shows the reading range measurement in the laboratory together

Figure 8. The reading range measurement in the laboratory and the measured reading range.

Figure 9. Application of the tag and the tag inserted into the end of the log.

The reading range from freshly cut wood is approximately 2.5 m at the European UHF RFID frequencies (865.6 - 867.6 MHz) in the laboratory measurements. For inserting the trans‐ ponder into the log, a simple tool or a manual applicator was developed. The applicator is made from an axe by replacing the blade with a holder for the transponder. Using this appli‐ cator, the tag is hit into the end of the log as shown in Figure 9. After some practice an oper‐ ator may mark up to 100 logs / hour with the first strike success rate of approximately 95 %. In addition to this manual application tool, a prototype for an automatic applicator for a for‐

harvester was developed [15].

**Figure 8.** The reading range measurement in the laboratory and the measured reading range.

with the measured reading range using TagFormance™ measurement device.

**Figure 7.** Simulation model of the transponder inside a log and the planar dipole antenna layout.

**4.2. RFID readers** 

estry harvester was developed [15].

The design of the transponder antenna was developed using electromagnetic simulations, laboratory tests and tests in production conditions in saw mills [14]. For electromagnetic simulations, Ansoft HFSS was used. The transponder readability is best when the tag is in the end of the log, as this part is usually exposed in the piles and on conveyor. If the transponder is in the side of the log it may be left under the log or covered by other logs and reading would have to be done through considerable thickness of wood and with the possibility of the tag being pressed against a metal surface. Figure 7 shows the simulator model of the transponder inside

Challenges and Possibilities of RFID in the Forest Industry

http://dx.doi.org/10.5772/54205

311

The planar dipole antenna was optimised for operation inside wet wood with tolerance for varying permittivity caused by varying moisture content in the wood. The electrical properties of wood were measured at UHF and the relative permittivity of the spruce was found to be of the order of 2.3 when fresh and 1.8 after kiln drying. Correspondingly, the loss tangent was 0.08 and 0.03 for fresh and dry spruce. When soaking wet, the relative permittivity of the wood may be even in the magnitude of 10. The final antenna design has the dimensions of 74 mm x 5 mm. Figure 8 shows the reading range measurement in the laboratory together

The reading range from freshly cut wood is approximately 2.5 m at the European UHF RFID frequencies (865.6 - 867.6 MHz) in the laboratory measurements. For inserting the transponder into the log, a simple tool or a manual applicator was developed. The applicator is made from an axe by replacing the blade with a holder for the transponder. Using this applicator, the tag is hit into the end of the log as shown in Figure 9. After some practice an operator may mark up to 100 logs / hour with the first strike success rate of approximately 95 %. In addition to this manual application tool, a prototype for an automatic applicator for a forestry

The desired high reliability in the wood tracing requires a good survivability of the trans‐ ponders, which can be only achieved by inserting the tags inside the log. For high readabili‐ ty in the different steps of the supply chain the best location for the tags is in the log end. The transponder size and shape have to be optimised for insertion into the log – several ap‐ proaches were investigated in the Indisputable Key –project [12] but a wedge-shaped trans‐ ponder that is punched into the wood was selected [13]. This transponder has the additional advantage of being difficult to remove from the log or to tamper with. The shape of the cas‐ ing with the inlay inside and the application method are illustrated in Figure 6.

**Figure 6.** Wedge-shaped transponder and its insertion into the wood.

To achieve good readability in the production conditions on the conveyors, a long reading range is needed. The transponder casing material is somewhat lossy (measured electrical loss tangent is ~0.03 at UHF) which limits the choice of possible transponder antennas to di‐ pole antennas. Wood is a natural material which is not isotropic or homogenous, and the moisture content varies greatly as the wood dries or gets soaked in rain after the tree is fell‐ ed. The moisture content affects greatly the permittivity and losses of the wood and thus the transponder antenna has to be designed to operate in the wood with varying electromagnet‐ ic properties. The moisture content may exceed 100 % of the dry material weight in fresh wood.

The design of the transponder antenna was developed using electromagnetic simulations, laboratory tests and tests in production conditions in saw mills [14]. For electromagnetic with the possibility of the tag being pressed against a metal surface. Figure 7 shows the simulator model of the transponder inside

simulations, Ansoft HFSS was used. The transponder readability is best when the tag is in the end of the log, as this part is usually exposed in the piles and on conveyor. If the trans‐ ponder is in the side of the log it may be left under the log or covered by other logs and reading would have to be done through considerable thickness of wood and with the possi‐ bility of the tag being pressed against a metal surface. Figure 7 shows the simulator model of the transponder inside the log and the basic layout of the planar dipole antenna inlay. The design of the transponder antenna was developed using electromagnetic simulations, laboratory tests and tests in production conditions in saw mills [14]. For electromagnetic simulations, Ansoft HFSS was used. The transponder readability is best when the tag is in the end of the log, as this part is usually exposed in the piles and on conveyor. If the transponder is in the side of the log it may be left under the log or covered by other logs and reading would have to be done through considerable thickness of wood and

the log and the basic layout of the planar dipole antenna inlay.

Figure 7. Simulation model of the transponder inside a log and the planar dipole antenna layout.

The planar dipole antenna was optimised for operation inside wet wood with tolerance for varying permittivity caused by varying **Figure 7.** Simulation model of the transponder inside a log and the planar dipole antenna layout.

the wood processing steps in the supply chain. The material has to be mechanically durable; sufficiently hard but not brittle and it may not absorb water. The transponders have to sur‐ vive several months in the logs. The material should have suitable electromagnetic proper‐ ties at UHF frequencies – ideally low loss and stable properties. In addition to the suitable chemical, mechanical and electrical properties the material has to be applicable for mass production of the transponders using common plastic fabrication techniques e.g. injection moulding. A suitable bio-composite material meeting these requirements is ARBOFORM® by Tecnaro GmbH [11] and it was selected as the transponder casing material. The ARBO‐ FORM® -material consists of lignin, natural fibres and processing aids. To facilitate the mass production, conventional plastic inlay with aluminium as the antenna pattern material was selected, as the amount of plastic in the inlays ending up into the pulping from the saw mill is negligible. Currently, paper inlays are also available for a non-plastic alternative. The

The desired high reliability in the wood tracing requires a good survivability of the trans‐ ponders, which can be only achieved by inserting the tags inside the log. For high readabili‐ ty in the different steps of the supply chain the best location for the tags is in the log end. The transponder size and shape have to be optimised for insertion into the log – several ap‐ proaches were investigated in the Indisputable Key –project [12] but a wedge-shaped trans‐ ponder that is punched into the wood was selected [13]. This transponder has the additional advantage of being difficult to remove from the log or to tamper with. The shape of the cas‐

To achieve good readability in the production conditions on the conveyors, a long reading range is needed. The transponder casing material is somewhat lossy (measured electrical loss tangent is ~0.03 at UHF) which limits the choice of possible transponder antennas to di‐ pole antennas. Wood is a natural material which is not isotropic or homogenous, and the moisture content varies greatly as the wood dries or gets soaked in rain after the tree is fell‐ ed. The moisture content affects greatly the permittivity and losses of the wood and thus the transponder antenna has to be designed to operate in the wood with varying electromagnet‐ ic properties. The moisture content may exceed 100 % of the dry material weight in fresh

The design of the transponder antenna was developed using electromagnetic simulations, laboratory tests and tests in production conditions in saw mills [14]. For electromagnetic

ing with the inlay inside and the application method are illustrated in Figure 6.

transponder is EPC Class 1 Generation 2 compatible.

310 Radio Frequency Identification from System to Applications

**Figure 6.** Wedge-shaped transponder and its insertion into the wood.

wood.

moisture content in the wood. The electrical properties of wood were measured at UHF and the relative permittivity of the spruce was found to be of the order of 2.3 when fresh and 1.8 after kiln drying. Correspondingly, the loss tangent was 0.08 and 0.03 for fresh and dry spruce. When soaking wet, the relative permittivity of the wood may be even in the magnitude of 10. The final antenna design has the dimensions of 74 mm x 5 mm. Figure 8 shows the reading range measurement in the laboratory together with the measured reading range using TagFormance™ measurement device. The planar dipole antenna was optimised for operation inside wet wood with tolerance for varying permittivity caused by varying moisture content in the wood. The electrical proper‐ ties of wood were measured at UHF and the relative permittivity of the spruce was found to be of the order of 2.3 when fresh and 1.8 after kiln drying. Correspondingly, the loss tangent was 0.08 and 0.03 for fresh and dry spruce. When soaking wet, the relative permittivity of the wood may be even in the magnitude of 10. The final antenna design has the dimensions of 74 mm x 5 mm. Figure 8 shows the reading range measurement in the laboratory together with the measured reading range using TagFormance™ measurement device.

**Figure 8.** The reading range measurement in the laboratory and the measured reading range.

Figure 9. Application of the tag and the tag inserted into the end of the log. **4.2. RFID readers**  The reading range from freshly cut wood is approximately 2.5 m at the European UHF RFID frequencies (865.6 - 867.6 MHz) in the laboratory measurements. For inserting the trans‐ ponder into the log, a simple tool or a manual applicator was developed. The applicator is made from an axe by replacing the blade with a holder for the transponder. Using this appli‐ cator, the tag is hit into the end of the log as shown in Figure 9. After some practice an oper‐ ator may mark up to 100 logs / hour with the first strike success rate of approximately 95 %. In addition to this manual application tool, a prototype for an automatic applicator for a for‐ estry harvester was developed [15].

production conditions – particularly to saw dust and wood splinters, and to the risk of im‐ pacts. To protect the readers and to facilitate their installation over the conveyor the com‐ mercial readers were enclosed into a robust aluminium casing with the antennas on the outside. The reader used was Sirit Infinity 510 with circularly polarised antennas. Figure 10 shows the reader installations in a saw mill in Sweden in the log sorting station and in the sawing. In the log sorting it was found that antennas in a frame around the conveyor gave

Challenges and Possibilities of RFID in the Forest Industry

http://dx.doi.org/10.5772/54205

313

RFID readers are used to read the transponders inside the log so that the logs can be identi‐ fied and information such as measurement data can be associated to the log or the associat‐ ed data can be retrieved. To identify the individual logs in addition to the reading of the transponder IDs, the ID-code has to be associated with the correct log on the conveyor. In the case of logs in the sawing this is relatively straight-forward as the logs are sawn top first so that the tags in the butt end of the log are always separated by at least the log length. This is based on the automatic applicator always inserting the transponder into the butt end of the log. The speed of the conveyor in sawing is reasonably low as well. In the log sorting the case is more challenging as in some saw mills the logs are not turned before the sorting and the transponders in the log ends can be very close to each other in adjacent log ends on the conveyor. To correctly identify the logs on the conveyor RFID positioning methods such as [16] could be used. In the Indisputable Key projects a simple method based on using the average reading time stamps from several antennas was used to determine the order of the transponders (and logs) on the moving conveyor). When the log separation was larger than about 1 m the logs could be identified reliably but some ambiguity in the log identification remained when the log separation was well below 1 m. The main reason for this was the difficult reading environment in the log sorting shown in Figure 9. There was a flat metal floor under the conveyor that causes reflections; the rapidly changing radio channel causes strong variation in the signal strength and variation of the position where the transponder is read on the conveyor. In the tests in other locations the log identification was significantly

more reliable reading of the tags than over the conveyor assembly.

**Figure 10.** RFID readers in the log sorting and sawing in a saw mill.

more reliable.

**Figure 9.** Application of the tag and the tag inserted into the end of the log.

#### **4.2. RFID readers**

Ideally, the traceability in the wood supply chain would reach from the forest all the way to the end user of the wood products - for full coverage of the supply chain RFID transponders would have to be read with RFID readers in every processing step shown in Figure 2. In the Indisputable Key project, the RFID based traceability was used in the round wood supply chain from harvesting in the forest to sawing at the saw mill. RFID readers were used in three processing steps: in the harvesting, at the log sorting and at the sawing where most of the information on the logs is collected and needed – hand-held readers can be used in other processing steps to supplement the fixed readers in the harvester and on the conveyor in the saw mills. The transponders and readers were compatible with the EPC Class 1 Gen 2 air interface standard.

Each reader installation site represents some unique challenges for the RFID reader and for its antennas. The RFID reader has to be able to read the transponders reliably from a practi‐ cal distance that depends on the location; for example on the conveyor in a saw mill the practical minimum distance from the reader antenna(s) to the transponder in the log is about 1 m as the thickness of the logs varies and sufficient space has to be left to accommo‐ date this variation. The forestry harvester represents the most challenging environment for the RFID readers in the wood supply chain; the reader is subjected to difficult electromag‐ netic and physical environment in outdoor conditions with rain, snow and ice, vibration, shocks, Nordic four season temperatures and also to occasional impacts. The developed pro‐ totype of a vibration and shock resistant RFID reader is described in [14, 15]. The RFID read‐ er features a robust impact resistant IP67 casing, adaptive RF front end for cancellation of reflections from large metal surface in the harvester head and EPC Global Reader Protocol v. 1.1 compatible interface over a CAN-bus to the harvester.

The reader installations at the saw mill were placed in the log sorting where the logs are first received and in the sawing. In these locations the RFID readers are subjected to industrial production conditions – particularly to saw dust and wood splinters, and to the risk of im‐ pacts. To protect the readers and to facilitate their installation over the conveyor the com‐ mercial readers were enclosed into a robust aluminium casing with the antennas on the outside. The reader used was Sirit Infinity 510 with circularly polarised antennas. Figure 10 shows the reader installations in a saw mill in Sweden in the log sorting station and in the sawing. In the log sorting it was found that antennas in a frame around the conveyor gave more reliable reading of the tags than over the conveyor assembly.

**Figure 10.** RFID readers in the log sorting and sawing in a saw mill.

**Figure 9.** Application of the tag and the tag inserted into the end of the log.

312 Radio Frequency Identification from System to Applications

1.1 compatible interface over a CAN-bus to the harvester.

Ideally, the traceability in the wood supply chain would reach from the forest all the way to the end user of the wood products - for full coverage of the supply chain RFID transponders would have to be read with RFID readers in every processing step shown in Figure 2. In the Indisputable Key project, the RFID based traceability was used in the round wood supply chain from harvesting in the forest to sawing at the saw mill. RFID readers were used in three processing steps: in the harvesting, at the log sorting and at the sawing where most of the information on the logs is collected and needed – hand-held readers can be used in other processing steps to supplement the fixed readers in the harvester and on the conveyor in the saw mills. The transponders and readers were compatible with the EPC Class 1 Gen 2 air

Each reader installation site represents some unique challenges for the RFID reader and for its antennas. The RFID reader has to be able to read the transponders reliably from a practi‐ cal distance that depends on the location; for example on the conveyor in a saw mill the practical minimum distance from the reader antenna(s) to the transponder in the log is about 1 m as the thickness of the logs varies and sufficient space has to be left to accommo‐ date this variation. The forestry harvester represents the most challenging environment for the RFID readers in the wood supply chain; the reader is subjected to difficult electromag‐ netic and physical environment in outdoor conditions with rain, snow and ice, vibration, shocks, Nordic four season temperatures and also to occasional impacts. The developed pro‐ totype of a vibration and shock resistant RFID reader is described in [14, 15]. The RFID read‐ er features a robust impact resistant IP67 casing, adaptive RF front end for cancellation of reflections from large metal surface in the harvester head and EPC Global Reader Protocol v.

The reader installations at the saw mill were placed in the log sorting where the logs are first received and in the sawing. In these locations the RFID readers are subjected to industrial

**4.2. RFID readers**

interface standard.

RFID readers are used to read the transponders inside the log so that the logs can be identi‐ fied and information such as measurement data can be associated to the log or the associat‐ ed data can be retrieved. To identify the individual logs in addition to the reading of the transponder IDs, the ID-code has to be associated with the correct log on the conveyor. In the case of logs in the sawing this is relatively straight-forward as the logs are sawn top first so that the tags in the butt end of the log are always separated by at least the log length. This is based on the automatic applicator always inserting the transponder into the butt end of the log. The speed of the conveyor in sawing is reasonably low as well. In the log sorting the case is more challenging as in some saw mills the logs are not turned before the sorting and the transponders in the log ends can be very close to each other in adjacent log ends on the conveyor. To correctly identify the logs on the conveyor RFID positioning methods such as [16] could be used. In the Indisputable Key projects a simple method based on using the average reading time stamps from several antennas was used to determine the order of the transponders (and logs) on the moving conveyor). When the log separation was larger than about 1 m the logs could be identified reliably but some ambiguity in the log identification remained when the log separation was well below 1 m. The main reason for this was the difficult reading environment in the log sorting shown in Figure 9. There was a flat metal floor under the conveyor that causes reflections; the rapidly changing radio channel causes strong variation in the signal strength and variation of the position where the transponder is read on the conveyor. In the tests in other locations the log identification was significantly more reliable.

#### **4.3. RFID system performance**

The RFID system performance in the traceability of round wood in the Nordic wood supply chain was tested in several trials in a saw mill in Sweden and in another saw mill in Finland [15]. In the tests the number of repeated transponder ID readings by the reader was found to be a good indicator of the reading reliability and means to compare reader set-ups. When the tag stays in the field of the reader, the reader keeps reading the ID of the tag repeatedly. With each reading event lasting about one milliseconds, the number of repeated readings in‐ dicates how long time the tag has been in the field of the reader. Table 1 shows an example of the observed average number of repeated readings in three tests – in Sweden 164 trans‐ ponders in 82 logs were run through the log sorting twice, and in a Finnish saw mill 143 test logs with transponders were sawn.

**Log marking Reader location Number of read tags**

Automatic in the

Automatic & manual, all logs for 26 Jan 2010

transponder reading rate.

UHF transponders in poles was excellent.

forest

**in the test**

Manual in the log yard Log sorting 218 207 95.0 %

**Table 2.** Examples of identification rates obtained in RFID tests in a Swedish saw mill.

**4.4. RFID use in other wood supply chains and processing steps**

Log sorting 285 268 94.0 %

Log sorting 812 754 92.9 %

The log identification rate was determined by synchronising the measuring time of the logs by 3D scanner in the log sorting and the RFID tag reading time. Due to the variation in the position of the transponder in the conveyor when it was read by the RFID reader located on the conveyor slightly after the 3D scanner, there was a time window for the time difference the reading timestamp and the 3D scanning timestamp. In the tests, there were also un‐ marked logs mixed with the RFID marked logs. When there was only one log inside this time window when the RFID tag was read, the log identification was considered successful. The achieved log identification rate was on average about 93 % in the log sorting at this saw mill and in other reading locations the log identification rate was practically the same as the

The promising results in the log identification using UHF RFID in the Nordic round wood supply chain created interest to test the capabilities of the RFID technology in tracing wood in other wood supply chains in the Indisputable Key project. Two other cases were investi‐ gated: wooden impregnated poles and sawn timber (boards). Impregnated poles are a prod‐ uct that has a supply chain similar to the round wood supply chain for sawn timber except that the wood used for poles has more stringent requirements and thus a higher value. Ad‐ ditionally, the impregnated wood is not used as a raw material for paper or any other prod‐ uct so there is no limitation for the materials to be used in the RFID transponders. The main challenge in the pole RFID marking is the impregnation process: the poles are impregnated with creosote in high temperatures exceeding +100°C and creosote is a powerful solvent of plastics. The tags are exposed to creosote for an extended time in these high temperatures. The impregnation of the poles destroys most commercial tags as well as the developed bio‐ degradable transponder. After some trials some special materials and high-temperature tol‐ erant commercial tags where found but their high prices made them not feasible for production use. Excluding the destruction of tags in the impregnation, the readability of the

The high readability of the RFID tags approaching 100 % caused the desire to try UHF RFID marking of sawn timber, i.e. boards, as the optical marking techniques can typically only reach at best up to 90-95 % readability of the markings in production conditions. The large

**Unique measurement results for the readings** **Log identification rate**

http://dx.doi.org/10.5772/54205

315

Challenges and Possibilities of RFID in the Forest Industry


**Table 1.** Reading tests with logs marked with UHF transponders.

Typically the obtained transponder reading rates exceeded 99 % in tests with some 200 logs. In practice, the maximum read rate is 300…600 times per second. As can be seen in Table 4.1 the deviation in the number of repeated ID readings (~120) is rather large compared to the average number of the repeats (180-190) in the log sorting at the saw mill in Sweden, where‐ as in the other reader location the deviation is smaller in relation to the average number of repeats (150 vs. 390) indicating a more reliable and consistent reading of the transponders. These results also show that for intact normally operating transponders the reading rate can be close to 100 %.

Tests with RFID marked logs were also carried out to determine the log identification rate in the log sorting station in the Swedish saw mill shown in Figure 9 (left-hand side). In this lo‐ cation, reflections from the metal floor caused ambiguity in the reading position on the con‐ veyor and the correct order of logs was unusually difficult to determine. Table 2 summarises the results from three tests where the log marking with RFID tags was carried out both in the forest and in the log yard at the saw mill using the manual applicator or the prototype of an automatic applicator in a forestry harvester.


**Table 2.** Examples of identification rates obtained in RFID tests in a Swedish saw mill.

**4.3. RFID system performance**

314 Radio Frequency Identification from System to Applications

logs with transponders were sawn.

**Table 1.** Reading tests with logs marked with UHF transponders.

an automatic applicator in a forestry harvester.

be close to 100 %.

The RFID system performance in the traceability of round wood in the Nordic wood supply chain was tested in several trials in a saw mill in Sweden and in another saw mill in Finland [15]. In the tests the number of repeated transponder ID readings by the reader was found to be a good indicator of the reading reliability and means to compare reader set-ups. When the tag stays in the field of the reader, the reader keeps reading the ID of the tag repeatedly. With each reading event lasting about one milliseconds, the number of repeated readings in‐ dicates how long time the tag has been in the field of the reader. Table 1 shows an example of the observed average number of repeated readings in three tests – in Sweden 164 trans‐ ponders in 82 logs were run through the log sorting twice, and in a Finnish saw mill 143 test

**Test Number of transponders Reading rate Number of repeated**

Log sorting test 1 164 100 % 190 120 Log sorting test 2 164 99.4 % 180 120 Sawing test 143 99.3 % 390 150

Typically the obtained transponder reading rates exceeded 99 % in tests with some 200 logs. In practice, the maximum read rate is 300…600 times per second. As can be seen in Table 4.1 the deviation in the number of repeated ID readings (~120) is rather large compared to the average number of the repeats (180-190) in the log sorting at the saw mill in Sweden, where‐ as in the other reader location the deviation is smaller in relation to the average number of repeats (150 vs. 390) indicating a more reliable and consistent reading of the transponders. These results also show that for intact normally operating transponders the reading rate can

Tests with RFID marked logs were also carried out to determine the log identification rate in the log sorting station in the Swedish saw mill shown in Figure 9 (left-hand side). In this lo‐ cation, reflections from the metal floor caused ambiguity in the reading position on the con‐ veyor and the correct order of logs was unusually difficult to determine. Table 2 summarises the results from three tests where the log marking with RFID tags was carried out both in the forest and in the log yard at the saw mill using the manual applicator or the prototype of

**readings per tag**

**Standard deviation of the repeats**

The log identification rate was determined by synchronising the measuring time of the logs by 3D scanner in the log sorting and the RFID tag reading time. Due to the variation in the position of the transponder in the conveyor when it was read by the RFID reader located on the conveyor slightly after the 3D scanner, there was a time window for the time difference the reading timestamp and the 3D scanning timestamp. In the tests, there were also un‐ marked logs mixed with the RFID marked logs. When there was only one log inside this time window when the RFID tag was read, the log identification was considered successful. The achieved log identification rate was on average about 93 % in the log sorting at this saw mill and in other reading locations the log identification rate was practically the same as the transponder reading rate.

#### **4.4. RFID use in other wood supply chains and processing steps**

The promising results in the log identification using UHF RFID in the Nordic round wood supply chain created interest to test the capabilities of the RFID technology in tracing wood in other wood supply chains in the Indisputable Key project. Two other cases were investi‐ gated: wooden impregnated poles and sawn timber (boards). Impregnated poles are a prod‐ uct that has a supply chain similar to the round wood supply chain for sawn timber except that the wood used for poles has more stringent requirements and thus a higher value. Ad‐ ditionally, the impregnated wood is not used as a raw material for paper or any other prod‐ uct so there is no limitation for the materials to be used in the RFID transponders. The main challenge in the pole RFID marking is the impregnation process: the poles are impregnated with creosote in high temperatures exceeding +100°C and creosote is a powerful solvent of plastics. The tags are exposed to creosote for an extended time in these high temperatures. The impregnation of the poles destroys most commercial tags as well as the developed bio‐ degradable transponder. After some trials some special materials and high-temperature tol‐ erant commercial tags where found but their high prices made them not feasible for production use. Excluding the destruction of tags in the impregnation, the readability of the UHF transponders in poles was excellent.

The high readability of the RFID tags approaching 100 % caused the desire to try UHF RFID marking of sawn timber, i.e. boards, as the optical marking techniques can typically only reach at best up to 90-95 % readability of the markings in production conditions. The large volumes of the boards sawn and the relative low value of the softwood boards excluded the use of cased transponders (hard tags) due to their price. Thus the only option was to experi‐ ment with label tags attached to the boards. The best readability with a label tag on the sur‐ face of fresh moist board immediately after sawing was achieved using an inlay indented for near metal applications with good performance in close proximity of detuning materials such as wood – e.g. UPM Raflatac Hammer. The achieved reading range was sufficient for board conveyors to ensure nearly 100 % readability where the reader antenna can be placed approximately 0.4 m away from the boards. However, the application of the label tags on the boards proved to be problematic. Different glues and stapling with plastic staples were tested but the transponder survival on boards in the saw mill in the processing steps from sawing to packing of the dried boards proved to be low - up to 30-40 % of the label tags at‐ tached to the boards after the sawing were lost before the packing. Thus the resulting tracea‐ bility of the boards would be too low for useful applications in the range of some 60 %.

sponsible for sending the event messages to the right subscribers. The Collaborative Messag‐ ing System is also responsible for authentication and authorization. The Traceability Services is responsible for storing the Traceability Data and presenting it to the users in cor‐

Challenges and Possibilities of RFID in the Forest Industry

http://dx.doi.org/10.5772/54205

317

The interfaces between architecture modules are specified to each message format:

when tag readers interact with upper levels is specified in [19].

**•** C1G2, UHF Class 1 Generation 2 Tag Interface standard specifies the interface between RFID readers and RFID tags. The specification describes the interactions between readers and tags and tag operating procedures and commands. The full specification can be

**•** RFID reads are individual reads of a RFID tag. The specification of the protocol used

rect format.

**Figure 11.** ICT System Architecture.

found from [18].

#### **4.5. ICT solution**

In the Indisputable Key project an ICT system solution was developed to handle the data storage and transfer to enable efficient utilisation of the collected information by different actors in the value chain.

The ICT System Architecture connects the enterprise business processes to the actual flow of objects. The architecture consist of tags to mark the individual objects, readers to observe the movements of tagged objects, reader data processor to interpret the raw RFID reads to basic observation events and an adapter to create the meaningful business events from the RFID events. The Traceability Services that provides the services to analyse and use the informa‐ tion and the Local ONS that provides the way of publishing the services to the other busi‐ ness partners and the users. Figure 11 presents the overall data flow of the architecture.

The ICT system architecture follows the guidelines set by the EPCGlobal architecture. The EPCglobal Network Architecture Framework is a collection of interrelated standards for hardware, software, and data interfaces, together with core services that are operated by EPCglobal and its delegates, all in service of a common goal of enhancing the supply chain through the use of Electronic Product Codes (EPCs).

Traceability Services Architecture extends the EPCglobal scope by offering the way to use other codes than EPCs and by providing Traceability Services. Traceability Services offers methods to monitor and optimize of the forestry wood supply chain, to research wood property correlations, and of course to trace the wood material throughout the supply chain. By tracing the wood object and processes used to manufacture the wood product the Tracea‐ bility Services offers the chain-of-custody and environmental product declaration for wood products.

The architecture comprises three modules: Adapter, Collaborative Messaging System and Traceability Services. Adapters are used to acquire traceability information from the proc‐ esses. The Adapters connect the observations of objects to the process data, generate events and send the events to the Messaging System. The Collaborative Messaging System is re‐ sponsible for sending the event messages to the right subscribers. The Collaborative Messag‐ ing System is also responsible for authentication and authorization. The Traceability Services is responsible for storing the Traceability Data and presenting it to the users in cor‐ rect format.

**Figure 11.** ICT System Architecture.

volumes of the boards sawn and the relative low value of the softwood boards excluded the use of cased transponders (hard tags) due to their price. Thus the only option was to experi‐ ment with label tags attached to the boards. The best readability with a label tag on the sur‐ face of fresh moist board immediately after sawing was achieved using an inlay indented for near metal applications with good performance in close proximity of detuning materials such as wood – e.g. UPM Raflatac Hammer. The achieved reading range was sufficient for board conveyors to ensure nearly 100 % readability where the reader antenna can be placed approximately 0.4 m away from the boards. However, the application of the label tags on the boards proved to be problematic. Different glues and stapling with plastic staples were tested but the transponder survival on boards in the saw mill in the processing steps from sawing to packing of the dried boards proved to be low - up to 30-40 % of the label tags at‐ tached to the boards after the sawing were lost before the packing. Thus the resulting tracea‐ bility of the boards would be too low for useful applications in the range of some 60 %.

In the Indisputable Key project an ICT system solution was developed to handle the data storage and transfer to enable efficient utilisation of the collected information by different

The ICT System Architecture connects the enterprise business processes to the actual flow of objects. The architecture consist of tags to mark the individual objects, readers to observe the movements of tagged objects, reader data processor to interpret the raw RFID reads to basic observation events and an adapter to create the meaningful business events from the RFID events. The Traceability Services that provides the services to analyse and use the informa‐ tion and the Local ONS that provides the way of publishing the services to the other busi‐ ness partners and the users. Figure 11 presents the overall data flow of the architecture.

The ICT system architecture follows the guidelines set by the EPCGlobal architecture. The EPCglobal Network Architecture Framework is a collection of interrelated standards for hardware, software, and data interfaces, together with core services that are operated by EPCglobal and its delegates, all in service of a common goal of enhancing the supply chain

Traceability Services Architecture extends the EPCglobal scope by offering the way to use other codes than EPCs and by providing Traceability Services. Traceability Services offers methods to monitor and optimize of the forestry wood supply chain, to research wood property correlations, and of course to trace the wood material throughout the supply chain. By tracing the wood object and processes used to manufacture the wood product the Tracea‐ bility Services offers the chain-of-custody and environmental product declaration for wood

The architecture comprises three modules: Adapter, Collaborative Messaging System and Traceability Services. Adapters are used to acquire traceability information from the proc‐ esses. The Adapters connect the observations of objects to the process data, generate events and send the events to the Messaging System. The Collaborative Messaging System is re‐

**4.5. ICT solution**

products.

actors in the value chain.

316 Radio Frequency Identification from System to Applications

through the use of Electronic Product Codes (EPCs).

The interfaces between architecture modules are specified to each message format:


**•** Event is specified to be one observation concerning an individual object. The interface used to transmit the event to the Adapter is EPC global's Filtering & Collection (ALE) In‐ terface, that specifies the delivery of event data to the upper roles. The event in this level could be "At location X in time Y the object with EPC was observed".

The Collaborative Messaging System is realized using publish-subscribe pattern. Publishsubscribe is an asynchronous pattern where publishers of events are not sending the events to predefined subscribers. Instead of sending the message to predefined subscriber, message is published with some topic and content. In forestry-wood production system each event must contain event providers ID, detected object ID and a time stamp. Event can also con‐ tain some measurement information. For example in log sorting the event can contain meas‐

Any defined Event provider is an event provider in traceability system. IAD event messages are published about events concerning IAD objects and process information events are pub‐ lished about information concerning processes that can't be focused to an individual object. Each IAD event message must contain id of an event provider, id of an object and an obser‐ vation time, which is the instant of time when the observation took place. An IAD event message can also contain measurement information about an observed object. For example in log reception station a log is measured with a 3-D scanner. These measurements are in‐

Send port 1

Message box

Send port n

IAD Event

IK Adapter

Challenges and Possibilities of RFID in the Forest Industry

http://dx.doi.org/10.5772/54205

319

IAD Event

Subscriber Service

urements that 3-D scanner read from an object.

cluded into an IAD event message.

IK Architecture

Subscriber

Traceability

Serv ci es

Messaging System

Event Prov di er

Publish an IAD Event (IK Adapter)

**Figure 13.** Collaborative Messaging System Data Flow.

IAD Event

Receive port


The Figure 12 presents the architecture when used across enterprises that do not use the same Collaborative Messaging Service. The data flow between enterprises can be realized by using the IAD events. Some application in Enterprise B can subscribe to the events produced in company A. Another connection point is ONS, for example end customers or parties not included into the production chain is to use ONS to look up for the service and use Services provided by the Traceability services to fetch the needed information. For example - custom‐ er can fetch a Chain-Of-Custody document or Environmental Product Declaration for object.

**Figure 12.** ICT system architecture across enterprises.

The centre of the ICT system architecture is a Collaborative Messaging System that is re‐ sponsible of transmitting the messages from publishers to correct subscribers. The Publish‐ ers are not aware of Subscribers and all the authorization and authentication is performed by the Messaging System.

The Collaborative Messaging System is realized using publish-subscribe pattern. Publishsubscribe is an asynchronous pattern where publishers of events are not sending the events to predefined subscribers. Instead of sending the message to predefined subscriber, message is published with some topic and content. In forestry-wood production system each event must contain event providers ID, detected object ID and a time stamp. Event can also con‐ tain some measurement information. For example in log sorting the event can contain meas‐ urements that 3-D scanner read from an object.

Any defined Event provider is an event provider in traceability system. IAD event messages are published about events concerning IAD objects and process information events are pub‐ lished about information concerning processes that can't be focused to an individual object. Each IAD event message must contain id of an event provider, id of an object and an obser‐ vation time, which is the instant of time when the observation took place. An IAD event message can also contain measurement information about an observed object. For example in log reception station a log is measured with a 3-D scanner. These measurements are in‐ cluded into an IAD event message.

**Figure 13.** Collaborative Messaging System Data Flow.

**•** Event is specified to be one observation concerning an individual object. The interface used to transmit the event to the Adapter is EPC global's Filtering & Collection (ALE) In‐ terface, that specifies the delivery of event data to the upper roles. The event in this level

**•** Application specific format is used to connect the business data to the object observations. For example IK Adapter receives a measurements made by 3-D scanner are received as flat-file. The IK Adapter then connects the measurement information to the event infor‐

The Figure 12 presents the architecture when used across enterprises that do not use the same Collaborative Messaging Service. The data flow between enterprises can be realized by using the IAD events. Some application in Enterprise B can subscribe to the events produced in company A. Another connection point is ONS, for example end customers or parties not included into the production chain is to use ONS to look up for the service and use Services provided by the Traceability services to fetch the needed information. For example - custom‐ er can fetch a Chain-Of-Custody document or Environmental Product Declaration for object.

Event providers (IK Adapter)

The centre of the ICT system architecture is a Collaborative Messaging System that is re‐ sponsible of transmitting the messages from publishers to correct subscribers. The Publish‐ ers are not aware of Subscribers and all the authorization and authentication is performed

Collaborative Messaging System IAD Event

Traceability Services

IAD Event

Local ONS

Enterprise A (or group of enterprises)

IAD Event

Messaging Layer Application Layer Physical Layer Management Layer Data Flow

**Enterprise Information Systems** Measurement applications (3-d scanner, moisture measurement application, ...) ERP applications (Invoicing application , production planning application, ...)

Business data

could be "At location X in time Y the object with EPC was observed".

**•** IAD Event is specified to be one event concerning an individual object.

mation it received from the RFID-reader.

318 Radio Frequency Identification from System to Applications

**Enterprise Information Systems** ERP applications (Invoicing application , production planning application, ...)

**Figure 12.** ICT system architecture across enterprises.

by the Messaging System.

IAD Event

ONS

Enterprise B

Subscription could be topic-based, content-based or a hybrid of these two. In a topic-based subscription a subscriber subscribes for an events published with some topic. In a content based subscription, subscriber receives an event if a content of the event matches to the con‐ straints defined by subscriber. Traceability architecture support hybrid of these two. IAD event providers publish events of a topic and subscribers can define content based subscrip‐ tions to one or more topics. For example - as illustrated in Figure 14Example IAD event data flow.

A harvester publishes two events with different topic:

**•** A LogHarvested event which contains the exact volume, quality and price information about log harvested

**Figure 15.** Supply chain steps with properties.

**5. Discussion**

new business models.

The purpose of Traceability Services is to act as a repository for item level traceability data and process level data and to provide services based on this information. The solution con‐ nects the steps of supply chain together and provides a common data model for the whole supply chain. The solution offers services for calculating of Environmental, Economical and Quality KPIs and analysis for the process data that are the basis for the KPI calculations.

Challenges and Possibilities of RFID in the Forest Industry

http://dx.doi.org/10.5772/54205

321

The forest industry represents some unique challenges to the traceability solutions – the da‐ ta is utilised by different actors in the value chain so that typically the information is pro‐ duced by one party and the information needs to be utilised by another party that may be outside the supply chain. The basis for the traceability and information utilisation is reliable and affordable identification of the wood. By identifying the wood material and items in the supply chain the associated information can be utilised by different parties. This enables new level of control of the wood conversion chain, tailored and specialised products, and

The main challenge in enabling the possibilities of the traceability in the forest industry has been the lack of a reliable and inexpensive means to identify automatically the logs and boards in various processing steps along the wood supply chain. The optical marking tech‐ niques such as printed markings offer the potential for very low costs but these methods struggle to reach better than 90-95 % success rate in the automatic identification of the wood items in industrial production conditions. The required identification success rate has to sig‐ nificantly exceed 99 % so that the information retrieval becomes a viable option for reacquir‐ ing the needed information, e.g. log dimensions. With 90-95 % identification success rate the risk for not being able to retrieve the needed data is some 10 times larger than what is gener‐ ally considered acceptable – the benefits of the traceability are quickly lost if the information

RFID technology offers the potential for near 100 % success rate in the identification of logs. The main challenge is the cost of the transponders – the acceptable cost for a transponder depends on the value of the wood material in question and on the expected savings and benefits to be obtained through the use of RFID. Currently the acceptable price for RFID var‐

cannot be retrieved for a significant percentage of the wood items.

**•** A HarvesterState event which contains information about harvester state (battery, fuel, position, etc…)

**Figure 14.** Example IAD event data flow.

There are three different subscribers for the event LogHarvested. Saw mill production plan‐ ner wants to preplan the production beforehand by knowing the quality and amount of logs that are about to arrive to the saw mill. Saw mill purchaser makes payment based on the log volumes harvested and Traceability application gathers the information for research. For the event HarvesterState there are two subscribers. Traceability application gathers data for re‐ search and Harvester company can monitor its harvester status.

By combining information throughout the supply chain the Traceability Services enables new methods of analyzing the wood material. The properties of wood object can be com‐ pared between different steps, see Figure 15 Supply chain steps with properties.

For example length in harvesting vs. length in log sorting. Another possibility is to analyze how some property affects some other property. For example, how an area of origin affects the board quality.

**Figure 15.** Supply chain steps with properties.

The purpose of Traceability Services is to act as a repository for item level traceability data and process level data and to provide services based on this information. The solution con‐ nects the steps of supply chain together and provides a common data model for the whole supply chain. The solution offers services for calculating of Environmental, Economical and Quality KPIs and analysis for the process data that are the basis for the KPI calculations.
