WO2022101366A1

WO2022101366A1 - System and method for human motion monitoring

Info

Publication number: WO2022101366A1
Application number: PCT/EP2021/081435
Authority: WO
Inventors: Kailash MANOHARA SELVAN; Thomas LABAT-CAMY
Original assignee: Spatialcortex technology Limited
Priority date: 2020-11-11
Filing date: 2021-11-11
Publication date: 2022-05-19
Also published as: GB202017785D0; WO2022101366A4; GB2600936A8; GB2600936A

Abstract

A human motion monitoring system for tracking a physical activity of interest of a human subject, comprises: a plurality of body posture measurement sensor units; at least one weight measurement sensor unit; a data hub; and a wireless communication link between the sensor units and the data hub. A method for human motion monitoring, comprises: capturing data from a plurality of body posture measurement sensor units and at least one weight measurement sensor unit, wherein the sensor units are worn by a subject being monitored; analysing the captured data; and designing an intervention for the subject based on the analysis of the captured data.

Description

SYSTEM AND METHOD FOR HUMAN MOTION MONITORING

FIELD OF INVENTION

The present invention relates to human motion monitoring

BACKGROUND OF THE INVENTION

There is a general desire to monitor human motion in numerous fields, including, workplace ergonomic comfort, sports, fitness training and physical rehabilitation.

Current human motion monitoring is either carried out visually which is subjective, or optical motion capture systems which are usually suited for only indoor tracking, or other wearable state-of-the art products which may not provide all the parameters to draw detailed bio-mechanical insights about the human motion or physical activity.

For the purpose of work-place ergonomics and manual handling at work-place, Work-related Musculoskeletal Disorders (WRMSD) which occurs due to incorrect manual handling is a leading source of injury at work-place, contributing over a third of all work-place injuries globally. Lack of consistent manual handling risk assessments is often cited as a key reason for occurrence of WRMSDs. Current risk assessment practices are subjective as they are based on visual inspections, are not continuous and may fail to capture all infield risk scenarios faced by the workers. The wearable ergonomic monitoring devices available today may only track posture of selected body parts like back and neck, hence may not be capable of tracking full body injury risks which maybe important for manual handling that involves the awkward posture of the full body, pushing, pulling, lifting, carrying, including others.

For the purpose of sports, fitness performance training, the wearable devices available currently including step counters, heart rate monitors and selected part posture monitors may fail to track the fullbody posture along with additional force experienced by the Musculo-skeletal system as a result of the physical activity, the understanding of which may be necessary to prevent any injuries and choose a suitable training program to achieve optimal performance.

For the purpose of physical rehabilitation, recovery from an injury or ailment, say gait rehabilitation, the wearable devices available currently may not be capable of tracking the lower limb posture along with the load bearing capability of individual feets in the process of recovery, the understanding of which may be necessary to choose a customized rehabilitation training schedule for optimal recovery. SUMMARY OF THE INVENTION

Systems (devices) and methods according to the invention are defined in the claims.

A human motion monitoring system for tracking a physical activity of interest that comprises a plurality of wearable sensor units of at least two types, more particularly, plurality of body posture measurement sensor units and at least one weight measurement sensor unit, a data hub, wireless communication between the sensor units and data hub. An associated method for data capture, data analysis and intervention is disclosed which may involve some of the following steps; capture of the data from multiple sensor units simultaneously from the disclosed device at every frame, creation of combined data packet at every frame, simultaneous calibration of multiple sensor units of the device, creation and analysis of a bio-mechanical model of the human, detection of abnormality by comparing to reference data sets, trigger a feedback when the analysis is carried out real time, a means that allows human to log any symptoms such as pain and discomfort experienced by the human while or after performing the physical activity of interest which may be applied to train machine learning models which may then be used as reference data as explained previously to detect future abnormalities.

The inventors have found that the invention has several advantages including the following listed,

Firstly, the device may provide comprehensive set of parameters to track both posture and additional forces or change in weight experienced by the human as a result of the physical activity;

Secondly, the device may be less intrusive to wearer while performing the physical activity of interest and may be suitable for outdoor and infield use; and

Thirdly, the device and the method may allow detection of abnormality while comparing to reference dataset and may provide real-time feedback when an abnormality is detected so that the human acts upon correcting the abnormality, thereby preventing injuries.

A human motion monitoring device for tracking a physical activity of interest that comprises a plurality of wearable sensor units of at least two types, more particularly, plurality of body posture measurement sensor units and at least one weight measurement sensor unit, a data hub, wireless communication between the sensor units and data hub.

Each of the plurality of body posture measurement sensor units according to the first aspect of the present invention may comprise a sensing unit, a microcontroller unit (MCU) based measurement circuitry, a radio-frequency (RF) unit for wireless communication, a battery and also an optional feedback unit, all enclosed within a housing.

Each of the plurality of body posture measurement sensor units according to the first aspect of the present invention may use sensing units either a or b below, a) An Inertial Measurement Unit (IMU) that comprises in one unit an accelerometer, a gyroscope, also possibly a magnetometer, and also possibly a barometer b) A flex sensor comprising a piezo-resistive element When a is used, the measurement circuity derives orientation of the sensing unit through a sensor fusion algorithm such as a Kalman filter. When b is used, the measurement circuity measures the amount of bending of the sensing unit.

Each of the plurality of body posture measurement sensor units according to the first aspect of the present invention may be attached to various parts of the human body including upper limbs, lower limbs, torso, hip, neck and head in any particular configuration . If the human motion activity of interest involves the whole body, body posture measurement sensor units may be attached to several parts of the human body including upper limbs, lower limbs, torso, hip, neck and head. If the human motion activity of interest involves only the lower body or the upper body, body posture measurement sensor units may be attached to parts that are relevant to the human motion activity of interest.

The body posture measurement sensor units according to the first aspect of the present invention may attach onto or become part of an article of clothing or garment.

A weight measurement sensor unit according to the first aspect of the present invention may comprise a sensing unit made up of plurality of strain gauges or pressure sensors embedded onto a sole that is insertable or attachable to an article of a footwear, a housing that encompasses a microcontroller unit (MCU) based measurement circuitry, a radio-frequency (RF) unit for wireless communication, a battery and also an optional feedback unit.

The data hub according to the first aspect of the present invention may be a wearable unit or a personal computer or a mobile device. When the data hub is a wearable unit, it may encompass a microprocessor or microcontroller for data-processing, a RF unit, a battery, a data storage unit and possibly a feedback unit.

The RF unit of the body posture measurement sensor units and weight measurement sensor units may allow two-way wireless communication with a data hub. The body posture measurement sensor units and weight measurement sensor units may send out the measured data when a request for data is received from the data hub. The RF unit of the data hub may have additional functionality to transfer data to a remote server or cloud

The feedback unit of the body posture measurement sensor units and weight measurement sensor units may include a suitable feedback mechanism such as haptic, vibration, sound and light. The feedback may be activated when a request for activation is received from the data hub.

According to the second aspect of the present invention there is provided a method for data capture, data analysis and intervention based on data analysis.

The data capture method according to the second aspect of the present invention may involve some of the following steps,

Capture of the data from multiple sensor units simultaneously at every frame, particularly, from body posture measurement sensor units and weight measurement sensor units defined in first aspect. • Creation of combined data packet at every frame that encapsulates the data from multiple sensor units, particularly, orientation and weight data from body posture measurement sensor units and weight measurement sensor units respectively. The data packets may have a structure; optional start identifier bytes at the beginning, bytes corresponding to the orientation data from individual body posture measurement sensor units, bytes corresponding to weight data of one foot or both feet from weight measurement sensor units in any order and ending with optional end identifier bytes. The data packets may be created by the data hub and may be used for any subsequent data handling such as wireless data transfer, data analysis and data storage.

• A simultaneous calibration method that calibrates multiple sensor units, particularly the body posture measurement sensor units and weight measurement sensor units as defined in first aspect. IMU based motion tracking systems usually require a series of pre-defined static poses such as stand still and T-pose and pre-defined dynamic poses such as turning or swinging to gather information to correct for initial biases and offsets due to attachment of the IMU units to the body part. The weight measurement units may also require an offset and possibly a gradient calibration at the start. The calibration method defined herein, may involve a set of pre-defined static poses and dynamic movements that provides data to concurrently calibrate both the body posture measurement sensor units and the weight measurement sensor units.

The data analysis method according to the second aspect of the present invention may involve some of the following steps,

• Creation of a bio-mechanical model of the human body, possibly through a link-segment representation of human body, in which, the orientation of various body parts such as, upper limbs, lower limbs, torso, hip, neck, head and shoulders, of the biomechanical model and the Ground Reaction Force (GRF) or absolute weight or weight lifted by the human represented in the biomechanical model is derived from the body posture measurement sensor units and weight measurement sensor units respectively as defined in the previous aspects.

• The bio-mechanical model of the human body herein that may be solved in a static or dynamic fashion, to calculate several bio-mechanical parameters including, absolute limb angles with reference to a set origin or reference planes, relative angle between limbs or joint angles, distances between joints, angular velocities and acceleration, joint forces and moments. Some of these parameters may be compared with reference data sets to assess Musculo-skeletal health, performance, infer risk of injury and detect abnormality. The reference data set may include findings from scientific experiments, empirical correlations published in scientific literature, guidelines from regulatory bodies, correlations developed from historic data, machine learnt models, any pre-defined limits or thresholds. The reference data set may vary based on the application of interest.

• The data analysis including comparisons and computations as described herein may be performed by the data hub using a processing unit, or alternatively by a remote server or computer which has a suitable software program that analyses the data as per the aspects disclosed here. Such comparison and computations may be performed real-time as the data is being collected or at a later time on the recorded data possibly stored in data storage of the data hub.

• The results of the data analysis including comparisons and computations as described herein may be presented to the human or wearer through a data visualization terminal like a mobile application, or web portal or suitable software running on a computer. The results of the data analysis including comparisons and computations as described herein may be presented further to relevant parties such as wearer’s supervisor, coach, team mate, colleague, occupational health specialist or health care provider. The raw data and, or outputs of the analysis may be stored in the remote server. The data visualization terminal may access the data stored in the remote server and may use it for data presentation including generating plots, summary tables and reports.

The intervention method based on data analysis according to the second aspect of the present invention may involve some of the following steps,

• When the bio-mechanical analysis and data analysis including comparisons and computations as described herein is carried out real-time, when an abnormality is detected a decision may be made to trigger a feedback. The feedback to the human may be provided through the feedback unit as defined in the first aspect. Other relevant parties such as a supervisor or a team member or an occupational healthcare provider may also be notified on the occurrence of an abnormality.

• The data visualization terminal may allow user to log any symptoms such as pain and discomfort experienced by the human while or after performing the motion activity of interest. This data may be stored in the remote server along with corresponding raw data and, or the outputs of the analysis. The remote server may store similar data from several humans over time. All or part of the data stored in the remote server may be used to derive empirical correlations or train machine learning models which may be used as reference data as explained previously to detect future abnormalities.

Additional sensors units such as grip sensors in the hand, emg sensors, muscle activity sensors and heart rate monitor may be used to improve the accuracy of the bio-mechanical model analysis. Additional sensor systems tracking other parameters such as environmental conditions and location relevant to the application of interest may also be used in conjunction.

According to the third aspect of the present invention there is provided a human motion monitoring device that is applied for monitoring, reporting and intervening Musculo-skeletal injury risk hazards faced by a worker carrying out manual handling tasks at work-place,

According to the fourth aspect of the present invention there is provided a human motion monitoring device that tracks Musculo-skeletal parameters relevant to sports and fitness including, symmetry, postural abnormalities, risk of injury, gait parameters, repetition and bodily strain. According to the fifth aspect of the present invention there is provided a human motion monitoring device that tracks Musculo-skeletal parameters relevant to rehabilitation from injury including extent of motion, load bearing capability, symmetry, postural abnormalities, risk of injury, gait parameters, repetition and bodily strain.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to better understand the present invention, the invention will now be described by example and embodiments with reference to the following drawings:

Figure 1 is a schematic of a human motion monitoring device comprising of a plurality of body posture measurement sensor units, at least one weight measurement sensor unit and a data hub according to the present invention;

Figure 2a is a block diagram of a schematic of a posture measurement sensor unit of Figure 1 ;

Figure 2b is a block diagram of a schematic of a weight measurement sensor unit of Figure 1 ;

Figure 2c is a block diagram of a schematic of a data hub of Figure 1 ;

Figure 3 is a schematic of a data packet structure;

Figure 4a is a schematic illustration of a combined calibration sequence;

Figure 4b is a schematic plot of the orientation data from a posture measurement sensor unit during the combined calibration sequence of Figure 4a;

Figure 4c is a schematic plot of the weight data from a weight measurement sensor units during the combined calibration sequence of Figure 4a;

Figure 4d is a schematic plot of a calibration curve for a weight measurement sensor unit from the combined calibration sequence of Figure 4a;

Figure 4e is a schematic illustration of a physical activity after the calibration sequence of Figure 4a;

Figure 4f is a schematic plot of the corrected orientation data from a posture measurement sensor unit during the physical activity of Figure 4e;

Figure 4g is a schematic plot of the calculated weight lifted data from the weight measurement sensor units during the physical activity of Figure 4e;

Figure 5a is a simplified schematic of a 2D bio-mechanical analysis model;

Figure 5b is a simplified schematic of a 2D bio-mechanical analysis model of Figure 5a applied to spinal force calculation;

Figure 6a is a schematic plot of the calculated weight lifted data from the weight measurement sensor units during a physical activity; Figure 6b is a schematic plot of the calculated back inclination angles from the back posture measurement sensor unit during the physical activity of Figure 6a;

Figure 6c is a schematic plot of the calculated weight lifted data from the weight measurement sensor units during a badly performed physical activity; Figure 6d is a schematic plot of the calculated back inclination angles from the back posture measurement sensor unit during the physical activity of Figure 6c;

Figure 7 is a schematic illustration of a machine learning model trained with the posture data, weight data and symptoms information;

Figure 8 is a schematic diagram illustrating a human motion monitoring device that is applied for monitoring, reporting and intervening Musculo-skeletal injury risk hazards faced by a worker carrying out manual handling tasks at work-place;

Figure 9 is a schematic diagram illustrating a human motion monitoring device that is applied for Musculo-skeletal performance monitoring of a sports or fitness activity; and

Figure 10 is a schematic diagram illustrating human motion monitoring device applied to Musculo- skeletal performance monitoring of a gait rehabilitation activity.

DETAILED DESCRIPTION

A human motion monitoring device according to the first aspect of the present invention, designated 10, as shown in Figure 1 comprises a plurality of wearable sensor units of at least two types, more particularly, plurality of body posture measurement sensor units 1 and at least one weight measurement sensor unit 2, a data hub 3.

As shown in Figure 1 , each of the plurality of body posture measurement sensor units 1 according to the first aspect of the present invention may be attached to various parts of the human body including upper limbs, lower limbs, torso and hip. As shown in Figure 1 , the weight measurement sensor unit 2 may be attached to one foot or both feet.

Each of the plurality of body posture measurement sensor units 1 according to the first aspect of the present invention may measure the orientation about its orthogonal axis (x, y and z). Each of the plurality of body posture measurement sensor units 1 may become part of an article of clothing or garment to measure orientation resulted due to the movement of the wearer.

As shown in Figure 2a, each of the plurality of body posture measurement sensor units 1 may comprise a sensing unit like a IMU 201 , MCU 202, a RF unit for wireless communication 203, a battery 204 and an optional feedback unit 205, all enclosed within a housing. An IMU comprises in one unit an accelerometer (like ADXL345), a gyroscope (like L3G4200D), and also possibly a magnetometer (like HMC5883L). An IMU can be used to measure the orientation usually through an optimal fusion of individual orientation estimates from accelerometer, gyroscope and possibly magnetometer. This is because the orientation estimated by integrating angular velocity measured by a gyroscope over time leads to accumulated errors, while the orientation estimated by earth’s gravitational and magnetic field measured by accelerometer and magnetometer respectively is subject to high levels of noise. A sensor fusion algorithm like Kalman filter is commonly used to resolve this issue. This can be performed at the individual body posture measurement sensor units by the MCU (like STM32F030R8T6). A suitable RF module (like Bluetooth or wifi) may be used for 2-way wireless data communication between individual body posture measurement sensor units and the data hub.

Each of the plurality of weight measurement sensor units 2 of Figure 1 according to the first aspect of the present invention may be attached or inserted to an article of a foot wear, like a sole to measure the absolute weight or additional weight lifted or GRF resulted due to the wearer and their movement and physical activity.

As shown in Figure 2b, each of the plurality of weight measurement sensor units 2 may comprise multiple sensing unit 207 embedded onto a sole 206, a signal treatment component 208, MCU 202, a RF unit for wireless communication 203, a battery 204 and an optional feedback unit 205, all enclosed within a housing. A strain gauge may be used as the sensing unit to measure the absolute weight or additional weight lifted or GRF resulted due to the wearer and their movement and physical activity. The strain gauges are usually connected into bridges and sensor treatment components like HX711 are used to amplify the signal before being read by a MCU (like STM32F030R8T6). A suitable RF module (Bluetooth, wifi) may be used for 2-way wireless data communication between individual body weight measurement sensor units and the data hub.

The data hub 3 of Figure 1 according to the first aspect of the present invention may be a wearable unit or a personal computer or a mobile device.

As shown in Figure 2c, data hub 3 may encompass a processing unit 209, data storage 210, a RF unit for wireless communication 203, a battery 204 and an optional feedback unit 205. A processing unit (like BCM2837 or STM32F030R8T6) may carry out the analysis on the data received from the body posture measurement units and weight measurement units real-time or store it in a data storage (like a removable memory card) for analysis later. A suitable RF module (like Bluetooth or wifi) may be used for 2-way wireless data communication, in which the body posture measurement sensor units and weight measurement sensor units may send out the measured data when a request for data is received from the data hub. The RF unit of the data hub may have additional functionality to transfer data to a remote server or cloud.

The feedback unit 205 of Figure 2a, 2b and 2c may provide a suitable sensory input such as haptic, vibration, sound and light to prompt or alert the wearer of a certain situation. The feedback unit may be a vibration transducer or haptic feedback or a loud speaker for sound feedback or a light source for visual feedback. The feedback may be activated when a request for activation is received from the data hub. The battery 204 of Figure 2a, 2b and 2c may be a power source (like NiMH, LiPo).

According to the second aspect of the present invention there is provided a method for data capture, data analysis and intervention.

Data capture method

• Capture of the data from multiple sensor units simultaneously at every frame , particularly, from body posture measurement sensor units 1 and weight measurement sensor units 2 of Figure 1 defined in first aspect.

At every frame in time, the data hub may request the posture measurement sensor units and weight measurement sensor units for data in any sequence and capture the received data into the particular frame.

• Creation of combined data packet at every frame, that encapsulates the data from multiple sensor units, particularly, orientation and weight data from body posture measurement sensor units 1 and weight measurement sensor units 2 respectively of Figure 1 .

At every frame the data received by the data hub may be formed into a combined data packet. As shown in Figure 3, the data packets may have a structure; optional start identifier bytes 301 at the beginning, bytes corresponding to orientation of individual body parts 303 from individual body posture measurement sensor units and bytes corresponding to weight data 304 of one foot or both feet from weight measurement sensor units in any order and ending with optional end identifier bytes 302. When the orientation of the individual body part is represented by angles about the orthogonal axes, each angle may be represented as 2 byte unsigned integers, with each angle normalized to cover the whole range of 2 unsigned integer bytes, making a total of 6 bytes per orientation of individual body parts 303 of Figure 3. Weight data from each of the feet 304 of Figure 3 may be represented as 4 byte unsigned integers, with weight value normalized to cover the whole range of 4 unsigned integer bytes. The overall size of the data packet of Figure 3 may depend on the number of body posture measurement sensor units and weight measurement sensor units attached to the wearer. The optional start and end identifiers, 301 and 302 respectively in Figure 3 may be predefined at the start of the data capture or may change at every frame of data capture. These data packets may be used for any subsequent data handling such as wireless data transfer, data analysis and data storage.

• A simultaneous calibration method that calibrates multiple sensor units, particularly the body posture measurement sensor units 1 and weight measurement sensor 2 units of Figure 1 as defined in first aspect.

IMU based motion tracking systems usually require a calibration step to gather information to correct for initial biases and offsets due to how posture measurement sensor units are attached to the body part. The weight measurement units may also require an offset and possibly a gradient calibration at the start for accurate weight measurements. A combined calibration method is explained here that may involve a set of pre-defined static poses and dynamic movements that provides data to concurrently calibrate the body posture measurement sensor units 1 and weight measurement sensor 2 units of Figure 1 as defined in first aspect.

As shown in Figure 4a, an example calibration sequence may have a series of pre-defined static and dynamic poses; stand still, T-pose, Turn pose and walking.

For an example body part, say right upper arm, that creates angle a when swung front-to-back, P when swung left-to-right and y when turned, as shown in Figure 4a, the initial raw orientation data without correction may not be aligned to the known body part orientation and the raw orientation data without correction may show all three angles changing although moved about a fixed axis as seen in Figure 4b. The orientation measured during the standstill pose may be used to generate a complex conjugate and multiplied onto the subsequent orientation data, so that subsequent orientation data is measured from the known initial orientation. Additional complex conjugates from moving about a known axis, like T-pose and turn pose may be derived to correct for offsets due to how the sensors are attached to the body.

The weight measurement sensor units 2 of Figure 1 require an offset and possibly a gradient calibration at the start for accurate weight measurements. As shown in Figure 4c, the uncorrected raw weight measurement, may not read the expected 50% of the body weight when the weight is distributed across 2 feets (during standstill, T-pose and Turn pose), may not read 0 Kg as expected when a feet is lifted in free air (during walking) and may not read the full body weight when the whole body weight is supported on one feet (during walking), instead, it might read Whm, WOm and Wm respectively. A calibration relationship or curve 401 of Figure 4d may be derived by mapping the measured and expected weights during the calibration sequence and may be used to estimate an actual weight from subsequent measured weights through interpolation, extrapolation of a polynomial fit from the calibration relationship or curve 401 . Further, the additional weight or force experienced by the wearer due to physical activity such as lifting a weight, pushing, pulling and impact may be extracted by subtracting the known body weight from the summated actual weight from both feets.

For an example physical activity, lifting and carrying a weight, as shown in Figure 4e, correcting the raw orientation and weight data as mentioned above may result in accurate orientation and weight estimates. As shown in Figure 4f, the corrected orientation data may show big variations in angle a only, as expected. As shown in Figure 4e, the additional weight WL calculated by the subtracting the known body weight from the estimated actual weight is only a finite positive value during the lifting and carrying phase, as expected.

Data analysis method

• Creation of a bio-mechanical model of the human body, possibly through a link-segment representation of human body, in which, the orientation of various body parts of the biomechanical model and the GRF or absolute weight or weight lifted by the human represented in the biomechanical model is derived from the body posture measurement sensor units 1 and weight measurement sensor units 2 respectively of Figure 1 as defined in the previous aspects. As shown in Figure 5a, the 2D bio-mechanical model 500 of the human body during a lifting phase of load 501 , has several body parts inclined to the horizontal plane and has a GRF 508 at the foot. According to the second aspect of the present invention, the inclination of the back, upper arm, lower arm, hip, upper leg and lower leg with the horizontal plane may be (90-aB) 503, (90-aUA) 502, (90-aLA) 505, (90-aH) 504, (90-aUL) 506 and (90-aLL) 507 respectively as seen in Figure 5a may be derived from body posture measurement sensor units 1 of Figure 1 as perthe first aspect. According to the second aspect of the present invention, the total weight or GRF on the feet 508 which includes the total body weight W and the weight of load 501 WL as seen in Figure 5a may be derived from one or both of the weight measurement sensor units 2 of Figure 1 as perthe first aspect.

• The bio-mechanical model of the human body herein that may be solved in a static or dynamic fashion, to calculate several bio-mechanical parameters including, absolute limb angles with reference to a set origin or reference planes, relative angle between limbs or joint angles, distances between joints, angular velocities and acceleration, joint forces and moments.

The bio-mechanical model 500 of Figure 5a may be applied to calculate reactive force of compression Fc and reactive shear force Fs across the L5/S1 disc 519 of Figure 5b, by solving the equilibrium equations of F_C0MP and F_SHEAR, 510 and 512 of Figure 5b respectively.

Equation

In equations 2 and 4, aB may be derived from 503 of Figure 5a, b 515 of Figure 5b may be the horizontal distance from lower back to the centre of gravity of the upper body 513 of Figure 5b, h 514 of Figure 5b may be the horizontal distance from lower back to load 501 of Figure 5a, E 517 of Figure 5b may the moment arm of muscle force F_MUSC 511 about the L5/S1 disc 519 of Figure 5b, W_UB may be the weight of the upper body 518 of Figure 5b and W_L 516 of Figure 5b may be the weight of the load lifted 501 of Figure 5a. All these variables may be readily available from the bio-mechanical model 500 as seen in Figure 5a and published anthropometric data from a reference human of features similar to the wearer. There by, the reactive force of compression Fc and reactive shear force Fs across the L5/S1 disc 519 of Figure 5b may be calculated.

Additional bio-mechanical analyses are further described in Don B. Chaffin, et al., “Occupational Biomechanics”, Wiley-lnterscience (2006), which is incorporated by refence in its entirety herein for everything that it teaches.

• Some of the computed parameters from the bio-mechanical model 500 of Figure 5a and 5b, may be compared with reference data sets to assess Musculo-skeletal health, performance, infer risk of injury and detect abnormality. The reference data set may include findings from scientific experiments, empirical correlations published in scientific literature, guidelines from regulatory bodies, correlations developed from historic data, machine learnt models, any predefined limits or thresholds. The reference data set may vary based on the application of interest.

Some examples of reference data sets are “Manual handling assessment charts (the MAC tool)” by Health and Safety Executive (HSE), UK and “Scientific Support Documentation forthe Revised 1991 NIOSH Lifting Equation: Technical Contract Reports” by National Institute for Occupational Safety and Health (NIOSH), USA, which are incorporated by refence in its entirety herein for everything that it teaches.

As seen in Figure 6b, the forward-backward inclination angle 502 and sideward inclination angle 503 of the back derived from the data from back posture measurement sensor unit 1 of Figure 1 in accordance with the previous aspects, lies within the safe zone 504, where the risk of injury may be low. Even during the lifting phase, where the additional weight on the wearer, derived from the data from weight measurement sensor unit 2 of Figure 1 in accordance with the previous aspects, is a finite positive number as seen in Figure 6a, the inclination angles of the back 502, 503 during the corresponding timeframe is still within the safe zone 504, which may suggest that the physical activity performed was correctly and safely carried out.

As seen in Figure 6c, the forward -backward inclination angle 506 and sideward inclination angle 507 of the back derived from the data from back posture measurement sensor unit 1 of Figure 1 in accordance with the previous aspects, exceeds the safe zone 504, thereby the risk of injury may be high. The risk of injury may be further exacerbated during the lifting phase, where the additional weight on the wearer, derived from the data from weight measurement sensor unit 2 of Figure 1 in accordance with the previous aspects, is a finite positive number as seen in Figure 6c, which might suggest that the physical activity performed was incorrectly or abnormally carried out.

This comparison and computations as per the previous aspects and explained through Figures 6a-6d, may serve as examples as to how the correctness and risk of injury or abnormality of physical activity performed by the wearer may be assessed quantitatively herein.

• Such comparison and computations may be performed by the data hub 3 of Figure 1 using the processing unit 209 of Figure 2c, or alternatively by a remote server or computer which has a suitable software program that analyses the data as per the aspects disclosed here. Such comparison and computations may be performed real-time as the data is being collected or at a later time on the recorded data possibly stored in data storage 210 of Figure 2c.

• The results of the data analysis including comparisons and computations as described herein may be presented to the wearer through a data visualization terminal like a mobile application, or web portal or suitable software running on a computer. The results of the data analysis including comparisons and computations as described herein may be presented further to relevant parties such as wearer’s supervisor, coach, team mate, colleague, occupational health specialist or health care provider. The raw data and, or outputs of the analysis may be stored in the remote server. The data visualization terminal may access the data stored in the remote server and may use it for data presentation including generating plots, summary tables and reports.

Intervention method

• When the data analysis including comparisons and computations as described herein is carried out real-time, if an abnormality is detected a decision may be made to trigger a feedback so that the wearer or human may act upon reducing the abnormality. As seen in Figure 6c and 6d, a suitable instance to trigger a feedback may be 505 as highlighted when the risk level of injury may be high. The feedback to the human may be provided through the feedback unit 205 of Figures 2a-2c as defined in the first aspect. Other relevant parties such as a supervisor, coach, team mate, colleague, occupational health specialist or health care provider may also be notified on the occurrence of an abnormality.

• Data visualization terminal like a mobile application, or web portal or suitable software running on a computer, may have additional functionality to allow user to log any symptoms such as pain and discomfort experienced by the human while or after performing the physical activity of interest. This data may be stored in the remote server along with corresponding raw data and, or the outputs of the analysis. The remote server may store similar data from several humans over time. All or part of the data stored in the remote server may be used to derive empirical correlations or train machine learning models which may be used as reference data as explained previously to detect future abnormalities. As shown in Figure 7, a simple machine learning model like an Artificial Neural Network (ANN) may be trained with the posture data 701 and weight data 702 and symptom information 703 as explained herein in accordance with the aspects explained here. Such a machine learnt model like 700 of Figure 7 may be used as reference dataset as explained previously to detect future abnormalities or likelihood of having a symptom or discomfort. Such a machine learnt model may be built to incorporate additional input parameters (such as data from additional sensors and human or wearer anthropometry) and additional performance indicators as outputs.

The human motion monitoring device 10 of Figure 1 , may incorporate additional sensors like grip sensors in the hand, emg sensors, muscle activity sensors and heart rate monitor may be used to improve the accuracy of the bio-mechanical model analysis. Additional sensor systems tracking other parameters such as environmental conditions and location relevant to the application of interest may also be used in conjunction.

The human motion monitoring device 10 can have several applications and hence have several embodiments are possible.

One such embodiment according to the third aspect of the present invention is a human motion monitoring device that is applied for monitoring, reporting and intervening Musculo-skeletal injury risk hazards faced by a worker carrying out manual handling tasks at work-place, designated 80 of Figure 8. As seen in Figure 8, the body posture measurement sensor units 1 of Figure 1 according to the first aspect of the present invention may be attached onto or integrate with the article of the Personal Protective Equipment including helmet 85, vest 81 , belt 82 and trouser 83 to measure the orientation of body parts including head, upper limbs, torso, hip and lower limbs while the weight measurement sensor units 2 of Figure 1 may attach to or become part of article of safety shoe 84 and the data hub 3 may be a wearable unit. Such an embodiment along with the method for data capture, data analysis and intervention according to the second aspect of the present invention may be used, firstly, to provide quantitative manual handling risk assessments as compared to current subjective visual risk assessments, secondly, to track injury risk scenarios faced by the workers infield over an extended period of time and thirdly, to provide real-time intervention when the injury risk expose faced by workers exceed safe limits.

One such further embodiment according to the fourth aspect of the present invention is a human motion monitoring device that tracks Musculo-skeletal parameters relevant to sports and fitness including, symmetry, postural abnormalities, risk of injury, gait parameters, repetition and bodily strain . As seen in Figure 9, such an embodiment designated 90 is applied to a human performing weight 91 lifting fitness performing training in which, the body posture measurement sensor units 1 of Figure 1 according to the first aspect of the present invention may be attached onto several articles sports clothing 92 to measure the orientation of body parts including upper limbs, torso, hip and lower limbs while the weight measurement sensor units 2 of Figure 1 may attach to or become part of article of training shoe 93 and the data hub 3 may be a mobile device. Such an embodiment along with the method for data capture, data analysis and intervention according to the second aspect of the present invention may be used for several purposes which include tracking progress against an optimal training schedule and to help train to achieve an optimal pre-defined lifting technique

One such further embodiment according to the fifth aspect of the present invention is a human motion monitoring device that tracks Musculo-skeletal parameters relevant to rehabilitation from injury including extent of motion, load bearing capability, symmetry, postural abnormalities, risk of injury, gait parameters, repetition and bodily strain. As seen in Figure 10, such an embodiment designated 100 is applied to a human gait rehabilitation in which, the body posture measurement sensor units 1 of Figure 1 according to the first aspect of the present invention may be attached onto lower limbs by attachments which may be straps 101 to measure the orientation of body selected body like parts lower limbs while the weight measurement sensor units 2 of Figure 1 may attach to or become part of article of foot wear 102 and the data hub 3 may be a laptop. Such an embodiment along with the method for data capture, data analysis and intervention according to the second aspect of the present invention may be used for several purposes including tracking progress against an optimal rehabilitation schedule.

Claims

1 . A human motion monitoring system for tracking a physical activity of interest of a human subject, comprising: a plurality of body posture measurement sensor units; at least one weight measurement sensor unit; a data hub; and a wireless communication link between the sensor units and the data hub.

2. A human motion monitoring system as claimed in Claim 1 , wherein each of the body posture measurement sensor units can be attached to a different body part of the subject and used to measure the inclination or orientation of the body part to which it is attached.

3. A human motion monitoring system as claimed in Claim 1 , wherein the sensing unit of each of the body posture measurement sensor units comprises: an Inertial Measurement Unit (IMU) that comprises, in one unit, an accelerometer and a gyroscope, optionally a magnetometer and optionally a barometer; or a flex sensor comprising a piezo-resistive element.

4. A human motion monitoring system as claimed in any preceding claim, wherein each body posture measurement sensor unit is attachable onto or is an integral part of an article of clothing or garment to be worn by the subject.

5. A human motion monitoring system as claimed in any preceding claim, the at least one weight measurement sensor unit is configured to measure the absolute weight, an additional weight, or a Ground Reaction Force (GRF) of the subject.

6. A human motion monitoring system as claimed in any preceding claim, wherein each of the body posture measurement sensor units and the at least one of weight measurement sensor unit comprises: a sensing unit, a microcontroller unit (MCU) based measurement circuitry, a radio-frequency (RF) unit for wireless communication, a battery, and optionally a feedback unit.

7. A human motion monitoring system as claimed in any preceding claim, wherein the at least one weight measurement sensor unit comprises one or more strain gauges, and/or one or more force resistive sensors, and/or one or more pressure sensors.

8. A human motion monitoring system as claimed in any preceding claim, wherein the at least one weight measurement sensor unit is embedded onto a sole that is insertable or attachable to an article of a footwear.

9. A human motion monitoring system as claimed in any preceding claim, wherein the data hub is a wearable unit, a personal computer, or a mobile device.

10. A human motion monitoring system as claimed in any preceding claim, wherein the data hub comprises: a microprocessor or microcontroller for data-processing, a RF unit, a battery, a data storage unit, and optionally a feedback unit.

1 1 . A human motion monitoring system as claimed in Claim 10, wherein the RF unit of the data hub has functionality to transfer data to a remote server or a cloud-based server.

12. A human motion monitoring system as claimed in any preceding claims, comprising a two-way the wireless communication link between the body posture measurement sensor units, the at least one weight measurement sensor unit, and the data hub.

13. A human motion monitoring system as claimed in any preceding claim, wherein the feedback unit comprises a haptic vibration transducer, and/or a sound source, and/or a light source.

14. A human motion monitoring system as claimed in any preceding claim, wherein the feedback unit is activatable on receipt of a request for activation from the data hub.

15. A method for human motion monitoring, comprising: capturing data from a plurality of body posture measurement sensor units and at least one weight measurement sensor unit, wherein the sensor units are worn by a subject being monitored; analysing the captured data; and designing an intervention for the subject based on the analysis of the captured data.

16. A method for human motion monitoring as claimed in Claim 15, wherein the data from plurality of wearable sensor units is captured in a series of data capture frames and data is captured simultaneously in every frame.

17. A method for human motion monitoring as claimed in Claim 16, wherein the data captured in every frame is combined into a data packet for every frame, wherein, the data packet has a structure comprising: optional start identifier bytes at the beginning of the packet, bytes corresponding to the orientation data from individual body posture measurement sensor units and bytes corresponding to weight data of one foot or both feet from weight measurement sensor units in any order; and optional end identifier bytes.

18. A method for human motion monitoring as claimed in Claim 16 or 17, wherein, the data packets are used for subsequent data handling including wireless data transfer, data analysis, and/or data storage.

19. A method for human motion monitoring as claimed in any of Claims 15 to 18, further comprising concurrently calibrating the plurality of body posture measurement sensor units and the at least one weight measurement sensor unit, through a set of pre-defined static poses and/or dynamic movements of the subject while wearing the sensors.

20. A method for human motion monitoring as claimed in any of Claims 15 to 19, further comprising using the data from the plurality of body posture measurement sensor units and the at least one weight measurement sensor unit to create a bio-mechanical analysis model of the human body which is solvable in a static or dynamic fashion.

21 . A method for human motion monitoring as claimed in Claim 20, wherein results of the analysis of the bio-mechanical analysis model are compared with reference data sets to assess Musculo-skeletal health, performance, infer risk of injury, and detect abnormality, wherein the reference data set includes findings from scientific experiments, empirical correlations published in scientific literature, guidelines from regulatory bodies, correlations developed from historic data, machine learnt models, and/or any pre-defined limits or thresholds.

22. A method for human motion monitoring as claimed in Claim 20 or 21 , wherein analysis of the bio-mechanical model and comparison with reference data sets is performed by the data hub using a processing unit, or alternatively by a remote server or computer which has a suitable software program, and is performed real-time as the data is being collected and/or at a later time on recorded data.

23. A method for human motion monitoring as claimed in any of Claims 20 to 22, wherein analysis of the bio-mechanical model and comparison with reference data sets is performed in real-time and when an abnormality is detected, a feedback is triggered to the subject or another person, and/or an alert is notified to other relevant parties.

24. A method for human motion monitoring as claimed in any of Claims 20 to 23, wherein analysis of the bio-mechanical model and comparison with reference data sets is presented to through a data visualization terminal comprising a mobile application, or a web portal, or suitable software running on a computer, and optionally wherein the data visualization terminal allows a user to log any symptoms experienced by the subject during or after performing the motion activity of interest.

25. A method for human motion monitoring as claimed in Claim 24, wherein symptom data along with data from the plurality of body posture measurement sensor units and the at least one weight measurement sensor unit, from one or more subjects over any duration of time is used to derive empirical correlations or train machine learning models which are used as reference data to detect abnormalities. A method for human motion monitoring as claimed any of claims 15 to 25, further comprising using data from additional sensors units such as hand grip sensors, emg sensors, muscle activity sensors, and heart rate monitors to improve the accuracy of the bio-mechanical model analysis. A human motion monitoring device and or method according to any previous claims is used for monitoring, reporting and intervening Musculo-skeletal injury risk hazards faced by a worker carrying out manual handling tasks at work-place, for tracking Musculo-skeletal parameters relevant to sports and fitness including, symmetry, postural abnormalities, risk of injury, gait parameters, repetition and bodily strain, or for tracking Musculo-skeletal parameters relevant to rehabilitation from injury including extent of motion, load bearing capability, symmetry, postural abnormalities, risk of injury, gait parameters, repetition and bodily strain.