NO20190557A1

NO20190557A1 - Fish biometrics

Info

Publication number: NO20190557A1
Application number: NO20190557A
Authority: NO
Inventors: Jakob Blomgren; Gert Andersson; Duncan Platt
Original assignee: Afc Analytics As
Priority date: 2019-04-26
Filing date: 2019-04-26
Publication date: 2020-10-27

Description

Technical Field

The present invention relates to a capsule based measurement of biometric parameters with magnetometer and magnetic tags. The system finds use in precision livestock farming.

Background of the Invention

Fish health is of major importance to the aquaculture industry. In order to ensure running the best possible business, both the survival rate of the fishes, their health and their quality need to be monitored. As such being able to measure fish biometrics is a critical tool to establish fish behaviour, fish welfare, fish size and thereby fish growth rate.

There is no single sensor solution to measure the parameters mentioned above.

Fish biometrics can be measured using subcutaneous capsules. However, the biometric parameters are limited to, e.g. temperature, heart-rate and pressure. The capsules are expensive and large, and therefore surgical implantation is required. The present-day capsule solutions work but are intended for research purposes and not for general application in the aquaculture industry to perform precision livestock farming.

Fish behaviour and fish growth are often measured using underwater camera systems. These have the disadvantage of being expensive. Furthermore, it is difficult to measure a large sample of fish, and also a problem to know which fish are being measured. Advancements with 3D camera, recognition algorithms and machine learning are being made to improve these systems, making them more accurate and relevant, thus providing better statistical sample data for the aquaculture industry. However, these systems will be limited to what can be observed by the camera at one time and will not be able to acquire the amount of statistical data that is achievable with the capsule-based system. Camera systems will also need to be used in conjunction with other biometric sensors to obtain the complete information necessary for establishing the necessary biometric

parameters.

Passive subcutaneous LF RFID fish tags exist today in very small capsule formats that are suitable for being injected into fish. These products have no sensors and their read range is very limited.

There exist active fish tags with hydroacoustic communication that can communicate over a distance of several tens of meters. These tags lack a sensor platform to provide biometric data and are too large to be injectable.

Summary of the Invention

It is therefore an object of the present invention, as it is stated in the set of claims, to solve the problems mentioned above.

The present invention describes a platform for the measurement of biometric parameters comprising a capsule based magnetometer sensor and magnetic tags.

Brief Description of the Drawings

The invention will now be described in detail in reference to the appended drawings, in which:

Figure 1 shows an embodiment of the invention: magnets are attached to the gills and the tail of a fish. A magnetometer sensor is placed in the abdominal cavity,

Figure 2 is a schematic drawing of a miniature subcutaneous capsule according to the invention,

Figure 3 shows an example calculation of the change in the magnetic field ΔBx for different gill or tail movement amplitudes Δx, assuming a typical remanence value for an NdFeB-magnet Br = 1.3 T and a magnet volume of V=0.5 cm<3>, and

Figure 4 shows an example calculation of the magnetic field Bz for different magnet to magnetometer separation, assuming a typical remanence value for an NdFeB-magnet Br = 1.3 T and a magnet volume of V=0.5 cm<3>.

Detailed Description

In the described invention, magnetometer sensor (6) and magnetic tags (1) are applied to extract information regarding fish (5) size, gill (4) and tail (3) motion, and thereby heart rate, and breathing. This information can be used for the extraction of fish (5) behaviour, welfare, and growth rate from one miniature sensor Integrated Circuit (IC) (6). The small size and low power consumption of such magnetic sensors (6) allows for miniaturisation of the capsule (2) with sensors (6) down to injectable sizes.

In a typical embodiment, magnetic tags (1) are attached to the fish (5) gills (4) and/or tail (3), whereas a capsule (2) comprising a magnetometer sensor, eventual other sensors such as but not limited to a power source, a microcontroller, a RFID, and a communication means, is injected in the fish (5) abdominal cavity. This capsule can optionally be covered with a suitable coating that is inert and does not trigger rejection from the fish.

The sensor (6) data can be communicated from the capsule (2) together with the fish (5) ID using a communication means, for example a hydroacoustic communication transducer or a radio signal communication means, thus providing the fish (5) farmer with the relevant biometric data and parameter information throughout the lifecycle of farmed fish (5). A hydroacoustic communication transducer is more suited to aquatic animals and a radio signal communication means is more suited to terrestrial animals.

A magnetometer can measure the magnetic field in one (1D), two (2D) or three (3D) directions. Such a magnetometer, inside the capsule (2) that is subsequently injected into the fish (5) abdominal cavity, can be used to detect the magnetic field at that point.

The application of permanent magnetic tags to the fish (5) gill (4) and tail (3) can be used to produce a time varying magnetic field, dependent on the gill (4) and tail (3) movement of the fish (5), that is detectable with a 3-axis magnetometer inside the capsule (2). This embodiment is illustrated in figure 1, and an embodiment of the capsule (2) is described in figure 2. The amplitude and frequency of detected magnetic field, corresponding to the fish (5) gill (4) and tail (3) motion, can thereafter be used as health, activity and size indicators.

Furthermore, the average magnetic field strength will provide an indication of the distance from the magnetometer to the magnetic tags, thereby providing a signal and information corresponding to the size of the fish (5).

The main benefit of this method is that the magnetic tag is passive (no need for batteries) and that the magnetic field pattern from the tag is unaffected by the body of the fish (5) or by the surrounding seawater.

The invention allows for the extraction of many biometric parameters with the help of innovative algorithms and signal processing from just one miniature sensor IC (6) to save on power consumption and reduce capsule (2) size.

This solution allows for a capsule (2) with sensor (6) platform to be miniaturised down to injectable sizes with a sensor (6) performance that exceeds anything on the market today.

The low cost of the capsule (2) sensor (6) platform solution means that many capsules (2) can be deployed at a fish farm, without risk to the fish (5) welfare or surgical implantation methods, in order to create a large identifiable sample of fish (5) biometric data that far exceeds what is achievable today using the established methods.

The last 10 years has seen the rapid development of magnetometers that are small, low-power, low-cost ICs developed mainly for compassing applications for mobile phones and/or the automotive market. These magnetometers are quite similar in shape and functionality and often include integrated 3-axis magnetometers, pre-amplifiers, microcontroller and SPI/I2C-interface interface.

Many suppliers offer a single IC component that combine a 3-axis magnetometer with a 3-axis MEMS accelerometer (6DOF) and some even include a 3-axis gyroscope (9DOF), see for instance LSM9DS from ST which is 3.5 x 3 x 1 mm in size or the eCompass from mCube.

The magnetic sensor (6) principle varies between the different manufacturers, but anisotropic magnetoresistance (AMR) and Hall effect are the most common. To compare the performance of these low-cost magnetometers is unfortunately often difficult, since the datasheets present the performance data in different ways and the data given is often incomplete. A typical candidate for this application is Memsic 5883MA with a magnetic rms-noise of Brms = 0.04 µT measured with an update rate of 100 Hz.

The Earth's magnetic field strength varies at different locations from BE = 30 μΤ to 60 μΤ, with a horizontal component pointing towards the magnetic north. The inclination of the magnetic field, measured as the angle downwards relative to the horizontal plane, varies from zero at the equator to /- 90° at the north/south pole. The horizontal component of the Earth's magnetic field is the component used for finding the heading angle and it varies in the range from BE_H = 10 μΤ to 30 μΤ. In Sweden and Norway, it is approximately 15 μΤ

Magnetic tags to measure gill and tail movement

The amplitude and frequency of the gill (4) and tail (3) motion can be used as health, activity and size indicators. The motion of the gill (4) or tail (3) can be measured by attaching a permanent magnet (1) to the gill (4) or tail (3) and sense the variation of the magnetic field with a 3-axis magnetometer placed in the abdominal cavity. The benefit of this method is that the magnet (1) is passive (no need for batteries) and that the magnetic field pattern is unaffected by the body of the fish (5) or by the surrounding seawater. It can be preferred that a tagged fish (5) only has a single magnet, either at the gill (4) or the tail (3), since it can be practically difficult to distinguish the magnetic fields from two magnets (1) with a single 3-axis magnetometer.

It is worth noting that if two magnetically tagged fishes (5) come so close together that the magnets (1) are separated by less than a few centimetres, the magnets (1) might stick together which might cause damage to the gill (4) or the tail (3) of the fish (5). The sticking force can be reduced by having a few millimetre-thick coatings around the magnets, which is anyway recommended in order to protect the magnet (1) from corrosion and to protect the skin of the fish (5).

In cartesian coordinates, with the origin in the centre of the magnet (1) and with the z-axis aligned with the magnetic moment μ, the magnetic field B far away compared to the size of the magnet (1) can be approximated with the field from a magnetic dipole:

Bx = 3μ0μ/4π * xz/r<5>

By = 3μ0μ/4π * yz/r<5>

Bz = μ0μ/4π * (3z<2 >- r<2>)/r<5 >

where r is the distance to the magnet, The magnetic moment can be calculated from the remanence field Br and the volume V of the magnet (1) using:

μ = V * Br/ μ0.

We are interested in the change of the magnetic field measured at the magnetometer when the magnet (1) moves back and forth with the gill (4) or tail (3). Moving the magnet (1) or the point of measurement is analogous, so we can use the above equations to calculate the B-field changes when the magnetometer is moved instead. As an example, we assume the magnetometer to be at (-Δχ,Ο,ζ) and it moves to (Δχ,Ο,ζ). From the equations above we get the following change in the magnetic field ΔΒ at the magnetometer:

ΔBX = 3VBr/4n * 2Δxz/r<5>

ΔBy = 0

ΔBZ = 0

Inserting a typical remanence value for an NdFeB-magnet Br = 1.3 T and a magnet volume of V=0.5 cm<3>, allows to plot the expected variation of the magnetic field ΔΒx for different gill (4) or tail (3) movement amplitudes Δx and magnet (1) to sensor (6) distances z.

From the plot of figure 3 we can see that the change in magnetic field will increase with the amplitude of the magnet (1) oscillation. The amplitude will probably be smaller for the gill (4) motion and larger for the tail (3). In the plot of figure 3, an estimated detection limit line has been inserted at 0.1 μΤ. We can conclude that a small motion, for example a gill (4) motion, can typically be detected up to 35 cm and a large motion, for example a tail (3) motion, typically up to 50 cm with current technology.

Magnetic tags to measure gill and tail signal strength to determine size and growth rate

The same arrangement as in the previous section, with a magnet (1) attached to either the gill (4) or the tail (3), can also be used to measure the size and growth rate of the fish (5). The detected oscillation amplitude and frequency of the magnet (1) motion can be used as an indirect indicator of the size of the fish (5). In addition, the magnitude of the magnetic field is directly linked to the separation between the magnet (1) and the magnetometer and hence a direct measure of the size of the fish (5) and, over time, the growth rate. If we again consider a magnet (1) pointing towards the magnetometer with a separation z, the resulting Bz-field will be Bz = VBr/(2πz<3>)(neglecting the oscillations due to the motion of the gill (4) and tail (3)). Using Br=1.3 T and a magnet volume of V=0.5 cm<3>, the resulting Bz-field is plotted in figure 4.

To measure a static magnetic field is more difficult than a harmonic signal. A static measurement will be sensitive to a zero-field offset in the magnetometer and any drift over time in the offset value. The Earth's magnetic field will also be present as strong background field, but as the fish (5) swims around randomly oriented in the Earth's field this contribution will average out over time. Hence, the static field (plus any magnetometer offset) can be determined as an average value measured over longer times. An alternative approach is to use two magnetometers as a gradiometer. The magnetometers are placed at different distances from the magnetic tag(s). The magnetometers will detect the same magnetic background from the Earth's magnetic field, but the difference in the measured field will originate from the magnetic tag(s). In the plot of figure 4 the estimated detection limit has been set to 1 μΤ, which is 10 times higher than before due to the offset and time average issues. Assuming these conditions, we can expect a detection range of 50 cm from the magnetometer to the magnet.

MEMS Accelerometer technology

Accelerometers can be made based on different principles and the requirements on, e.g., performance, cost, and size results in different optimal solutions. For the present invention, the extreme constraints on size/weight and current consumption imply that only MEMS accelerometers are possible to use. This kind of accelerometer is made in silicon and benefits from the fact that the fabrication can use the mass-production facilities for electronic circuits. Although the processing methods that can be used is limited to etching, deposition, and similar, a surprising richness of functionality can be achieved in MEMS components.

A MEMS accelerometer typically has a suspended proof mass which is connected to the inside of the accelerometer capsule. This causes the proof mass to deflect in proportion to the capsule acceleration, and this deflection is measured using, e.g., piezoresistive or capacitive detection. It should be noticed that the simplest and smallest design possible is limited to one-dimensional (1D) deflection. As acceleration is a 3D quantity, a complete measurement needs a 3D accelerometer. This requires a more sophisticated design or the combined use of three 1D accelerometers.

An example of an extremely small MEMS accelerometer is the MC3672 accelerometer from mCube6. This component is tiny (1.1 mm x 1.3 mm) and needs only 0.9 µA supply current at 25 Hz. It comes in different ranges and has digital output (I2C, SPI). The sample rate is configurable from 14 to 1300 samples per second. The noise is given as 6.5 mg AMS and the bias stability is given as ±40 mg.

MEMS Gyro technology

A rigid object has six degrees of freedom (DOFs), of which three describe the position and three describe the attitude/orientation. The output of a 3D accelerometer is related to the position, as the acceleration is obtained by differentiating the position with respect to time two times. The change of the attitude with time, on the other hand, is described by the 3D angular rate (I.e., rotation) of the object. This can be measured using a 3D gyroscope.

Also, this kind of device comes in very different implementations. For example, optical gyros are common in high-performance applications. For the current invention, the choice is limited to MEMS components. However, as opposed to accelerometers, a gyroscope must be actively driven. The reason is that the component must be made to vibrate as this, when the component is subject to a rotation, will cause a Coriolis force and a measurable deflection proportional to the rotation rate.

The fact that active excitation is needed typically results in a larger size and driving current that for an accelerometer. As for the accelerometer case, a straightforward design is 1D, and the combination of three devices is typically needed to obtain the angular rate vector. (It may be possible to obtain useful information from a 1D gyro, e.g., the fish rotation due to the tail beat. This, however, requires alignment of the capsule within the fish and this complicates the insertion considerably.) Fish monitoring using accelerometer and gyroscope signals

Before discussing the possible use of the data, it is important to remember that also an accelerometer at rest gives a non-zero output. This is because the Earth's gravity gives rise to an (apparent) acceleration equivalent to what would result from an upward acceleration of 1 g. As acceleration is a vector quantity, a rotation will shift this signal to another axis (assuming a 3D accelerometer). To this is added the acceleration of the object and this makes it useful to divide the acceleration signal into the low (LF) and high frequency (HF) content. The limit between these parts needs can be reasonably defined at for example 1 Hz.

The LF part can be obtained by filtering or averaging and will provide information about the roll and pitch of the fish, i.e., about two out of the three attitude DOFs. However, the heading remains unknown. This is intuitively clear as knowledge about the up direction does not tell us where north is. In this context, it is worth pointing out that by combining accelerometer and magnetometer data, the (complete) attitude of the sensor can be determined. This information is, however, only available when the object is accelerating/rotating slowly enough to allow a steady LF component.

The LF component therefore can act as an inclinometer. This has at least two possible applications. First, on a long timescale, the average attitude of the fish may allow conclusions to be drawn about the fish health status as, same illnesses may affect the ability of the fish to maintain a levelled attitude. Second, on a timescale over minutes, the behaviour during feeding may be monitored. As found by Føre et al., salmon swims vertically to reach the pellets during feeding and this should cause a clear change of the attitude. As further reported, it seems possible to relate the swimming behaviour to the hunger of the salmon. Thus, by monitoring the salmon attitude, the feeding could be better adapted to the salmon interest in further feeding.

The HF part of the signal contains the faster dynamics caused either by acceleration or rotation that shifts the gravity signal between the axes. One parameter that can be monitored using this signal is the fish activity. To do this the average magnitude over a suitable time interval of the HF part could be used. This would include the motion pattern of the fish due to fin movements, including the tail beat. It is reasonable to assume that the activity found in this way would correlate with the fish metabolic rate.

A further interesting use of the HF part is as an indication of the fish size. As explained by Sato et al., there is reason to believe that when an animal grows larger, the frequency of the limb movement will decrease. Thus, if the tail beat of the salmon is measured during activity, then a mean frequency of the HF part should correlate with fish size. This could be obtained by finding the acceleration within a number of frequency bands and monitoring how this "size fingerprint" develops over time. Although individual measurements should be expected to be noisy, the long-term evolution should be possible to smooth by filtering.

Like the accelerometers, also the gyros would have a background signal. This is due to the Earth's rotation around its axis, but the difference is that this signal is much smaller than typical fish rotations and not the other way around, as is the case for the fish acceleration compared to gravity. Gyro data allows accurate tracking of the attitude, although supporting data is necessary to avoid long-term error accumulation. This supporting data is available from the inclinometer and magnetometer.

In short, it can be said that gyro data would complement all the above measurements and combining it with accelerometer data would make the identification of interesting parameters easier. As an example, measuring the tail beat frequency (TBF) may be significantly easier if the rotation around a vertical axis is available.

The sampling rate with which to perform these measurements remains to be determined. For example, more information about the TBF is necessary to make an informed decision. However, it seems likely that there is not much frequency content above 10Hz, and it may be possible to use a sampling rate of, say, 25 samples per second. Further questions relate to when to measure as the continuous operation and processing of acceleration data is likely to be too power consuming. A method for triggering the measurement from the outside would be an alternative, which would make it possible, e.g., to start the measurement on demand during feeding.

Temperature sensor technology

Temperature sensor inside the capsule's microcontroller can achieve a relatively good performance. Companies such as NXP have RFID single chip solutions (e.g. NHS3100 D) with inbuilt temperature sensors that are accurate to ±0.3 °C, over a limited temperature range between 0 °C and 40 °C. Such performance can be suitable for this application. The temperature information is of interest to understand fish behaviour. lf particularly accurate temperature sensing is required, then there exist chips scale package (< 1 mm<3>) solutions that could be applied that have a performance typically better than 0.05 °C, over a limited temperature range. Such a component would be suitable for capsule integration.

MEMS Microphone technology

MEMS microphones are an interesting alternative to other types of microphones due to their robust structure and small size. As an example of a compact design, there is the SPV1840LR5H-B provided by Knowles Electronics. The size of this component is 2.75 mm x 1.85 mm x 0.90 mm and the supply current is < 60 µA. The sensitivity is -38 dBV/Pa. This sensitivity is however frequency dependent and drops by 3 dB just below 100Hz. As a heartbeat spectrum has almost all of its spectral content below 100Hz, it is necessary to evaluate if the reduced sensitivity at low frequencies is acceptable.

The microphone output can be sampled directly, but due to the low signal level it may be necessary to amplify the signal in order to make it compatible with the ADC range. The sampling rate must be set according to the spectrum width.

There are also microphones that have a digital output, which would elimmate the need for an amplifier and an AOC. One example is the SPH0641LM4H-1 provided by Knowles Electronics. This component requires about ten times higher supply current and has a single bit pulse-density modulation (PDM) output. In this representation, an analogue signal level is, as the name implies, represented by the density of ones in a high-speed digital bit stream. For our purposes, although this reduces the need for analogue components, this is not necessarily a better alternative as the data rate of the bit stream is high.

MEMS Microphone to measure heartbeat

Using a microphone to measure the pulse is similar to how human heartbeats are normally detected with a stethoscope. The fish is living in a quiet environment and since the capsule is internal to the fish, there is good reason to believe that an acoustic signal should be sufficient to be able to identify the pulse clearly. Given the high sampling rate and the amount of processing needed, it is reasonable to assume that the pulse monitoring must be performed intermittently. (This is true also for the approaches in the following sections, although they may be able to use a lower sampling frequency.) The acoustic data would be sampled and stored during a selected time interval, and the pulse rate would then be extracted using either frequency analysis methods or some more advanced method that uses knowledge about the approximate shape of a heartbeat signal.

The selection of the placement of the microphone leads to one important choice. The microphone can either (i) be placed within the hermetically sealed capsule or (ii) be mounted with a more direct contact to the environment. The traditional utilization, microphones are exposed to the outside environment, air. However, in a harsh outside environment, water or other liquids may enter the microphone cavity, affecting the microphone performance and sound quality. Liquid entrance can also permanently damage the microphone. Thus, the microphone has to be covered by a membrane. The last alternative however leads to questions about the biocompatibility of the microphone material. The first alternative is much simpler from a robustness point of view, but will reduce the signal as the acoustic wave needs to pass the capsule wall.

MEMS Accelerometer to measure heartbeat

Using the accelerometer data to measure the heartbeats is another option. It is fully possible that the capsule is in close and constant contact with large vessels etc. in the abdominal cavity of the fish. In addition to the acoustic coupling this would then give rise to accelerations. The advantage of this approach is that the accelerometer is useful also for other purposes. The acceleration within the heart tissue is large (on a 1 g scale) and would give a very clear signal. If the capsule is in direct contact with a large vessel, there is also reason to believe that the method could work well. It should also be remembered that the total acceleration contains different signal originating from all kinds of motion.

ECG electrodes to measure heartbeat

Sophisticated monitoring of human hearts is traditionally done using electrocardiography. For humans, the measurement is performed by placing (about ten) electrodes on the skin and recording the voltages that result from the activity of the heart muscle cells. This allows measurements at different angles relative to the heart, but for the present purposes this is unnecessarily sophisticated. An approach to measure heartbeats has been to place electrodes on the outside of the capsule. A burst of data is recorded, and the heart rate is calculated together with a quality index of the recording. Either this is logged in a Data Storage Tag or the result is obtained via the AF telemetry system. The choice here is between performing calculations onboard and to communicate the larger raw data set. Both of these options have an impact on power consumption.

The advantage with this method is that it can be expected to work equally well as previous system, but it comes at the cost of introducing external electrodes and the necessary wiring through the capsule wall. These electrodes should be placed as far as possible apart in order to maximize the potential difference. The analogue signal likely needs amplification before it can be sampled.

MEMS pressure sensor technology to measure depth

A pressure sensor measures the same quantity as a microphone. However, microphones measure pressure variations, with a typical frequency of 1 kHz, at very low pressures, often well below 1 Pa, while pressure sensors suitable· for this purpose need to have a range of about 1 MPa and needs only a very low sampling rate. The pressure sensor must be mounted with its open end facing the volume in which the pressure is to be monitored. This makes is necessary to either mount the entire sensor outside the capsule or to design the capsule around the component such that the sensor opening is on the outside.

One example component with a very small size and current consumption is the gelfilled MS5837-30BA from TE Connectivity. This device is intended for wearable underwater use, e.g., in dive watches. It is 3.3 mm x 3.3 mm x 2.75 mm large and has a metal and ceramic package. Unfortunately, the ceramic part is quadratic, with the soldering parts in the corners. The range is 30 bar (approximately 300 m water depth) with a configurable resolution with a best value of 0.2 mbar (2 mm). It has an internal ADC and digital I2C output. The current consumption is between 0.6 µA and 20 µA depending on the operating condition. However, also with the lowest current consumption, the resolution is 16 mm, which should be sufficient for this purpose. The design is intended to use a 1.8 mm O-ring as the sealing component. A high-resolution output from an internal temperature sensor is also available. It is stated that the absolute accuracy is ±2 °C but it is not impossible that an on-site calibration can compensate for this. We also note that the placement of the temperature sensor in the pressure sensor should lead to a good thermal contact to the environment. We conclude that from the information available for this component, it is clear that the size and current consumption is low enough to make the inclusion of a pressure sensor a viable alternative.

Even though the concept of the invention has been described for use in aquaculture, in the field of fish farming, its potential is much wider. One could use this invention to measure biometrics in a wide range of individuals: humans, any animal, preferably suitable for precision livestock farming for animal husbandry, preferably any aquatic animals, especially fishes, more preferably any animal being suitable for aquaculture, especially fishes.

Claims

1. A system for measuring biometric parameters in an individual,

c h a r a c t e r i z e d i n t h a t the system comprises a capsule (2) adapted to be injected in the individual’s body, said capsule (2) including a power source, a microcontroller, a communication means and a magnetometer (6);

said system further including at least one magnetic tag (1) attached to the individual.

2. A system according to claim 1, wherein the communication means is a hydroacoustic communication transducer or a radio signal communication means.

3. A system according to claim 1, wherein the capsule (2) further includes an accelerometer, and/or a gyro, and/or a temperature sensor, and/or a microphone, and/or a pressure sensor, and/or a RFID tag or any combination thereof.

4. A system according to any of the previous claims, wherein the at least one magnetic tag (1) is attached in a permanent manner.

5. A system according to any of the previous claims, wherein the at least one magnetic tag (1) are permanent magnets or any magnetic material that produces a magnetic field strength detectable by the magnetometer (6).

6. A system according to any of the previous claims, wherein the magnetometer (6) is a 1, 2, or 3-axis magnetometer, preferably a 3-axis magnetometer.

7. A system according to any of the previous claims, wherein the magnetometer (6) is combined with a 3-axis MEMS accelerometer and/or a 3-axis gyroscope.

8. A system according to any of the previous claims, wherein the at least one magnetic tag (1) is coated with an inert material.

9. A system according to any of the previous claims, wherein the individual is a human or an animal, preferably an aquatic animal, most preferably a fish.

10. A system according to any of the previous claims, wherein the individual is a fish (5), and the at least one magnetic tag (1) is attached to the fish’s gills (4) and/or to the fish’s tail (3).

11. A system according to any of the previous claims, wherein the system is adapted to obtain biometric data regarding motion and size which subsequently is correlated and converted to health parameters such as heart-rate, breathing, growth rate, and behavioural parameters such as stress.

12. The use of a biometric measurement system according to any of the previous claims for precision livestock farming for animal husbandry.

13. The use of a biometric measurement system according to claim 12 where precision livestock farming for animal husbandry is aquaculture, more preferably fish farming.