Modeling differences in the vibration response characteristics of the human body

Mathematical models may provide a useful tool for the development and evaluation of seating systems for vibration mitigation. A five-degree-of-freedom (DOF) model was formulated based on the measured driving-point impedance and transmissibilities of major anatomical structures contributing to the observed resonance behaviors. The model was limited in its ability to simulate differences observed in the resonance behaviors of a broader population and was unable to simulate the multiple resonances observed in the thigh. This paper describes the effectiveness of a modified five DOF model in simulating the major resonance behaviors observed in the population using representative data from a 56 kg female and 75 kg male. In addition, the model was also evaluated for its ability to predict the effects of selected seat cushions. The modified lumped-parameter model improved the peak chest and spine transmissibility simulations. The model was effective in simulating both the lower impedance peak observed in the primary resonance region (4-8 Hz) and the prevalent impedance peak observed in the second resonance region (7-10 Hz) in the smaller subjects. However, the model was not effective in predicting the dampening observed in the second resonance peak with the use of cushions. Redistribution of the model coefficients for the legs and the consideration of coupling between the legs and other anatomical structures may further improve the ability of the lumped-parameter model to predict the effects of seating systems on vibration transmission in the human body.     

Impedance response characteristics of the primate Mucaca mulatta exposed to seated whole-body gz vibration.

A mathematical model was used to quantify and describe the variability in the mechanical impedance response of the Rhesus monkey subjected to vibrations in the range 3-20 Hz at 0.5 g peak acceleration. Due to the similarities in response, a two-mass, one-degree-of-freedom (DOF) model was selected and the associated mechanical parameters determined using a nonlinear least-squares optimization program. For the six tests conducted on each of the four subjects, appreciable parameter variations were observed within a subject; however, the majority of the mean parameter values among different subjects and among the repeated tests on the population were within +/- 1 S.D. of each other. Significant differences were observed in the stiffness coefficient and the total mass among different subjects, and in the mass ratio (between inert and sprung masses) among the repeated tests. Variations in the profile shapes following resonance were described and limited by changes in the mass ratio and the damping factor. Higher mass ratios (greater than 1.0) were associated with lower damping factors (less than 0.50). The impedance response beyond resonance approached the response described by the impedance of the inert mass and the damper elements of the model combined in parallel, and supported the assumption that the lower torso was rigidly attached to the seat. Physically, the reactive force produced by the upper torso increasingly diminished following resonance, due to the load transmission/attenuation characteristics of the spinal structures at 0.5 g peak acceleration. The impedance measured at the seat becomes dominated by the transmitted damping force associated with the spine and the force generated by the rigid lower-torso mass.     

Non-linear characteristics in the dynamic responses of seated subjects exposed to vertical whole-body vibration

The effect of the magnitude of vertical vibration on the dynamic response of the seated human body has been investigated. Eight male subjects were exposed to random vibration in the 0.5 to 20 Hz frequency range at five magnitudes: 0.125, 0.25, 0.5, 1.0 and 2.0 ms(-2) r.m.s. The dynamic responses of the body were measured at eight locations: at the first, fifth, and tenth thoracic vertebrae (T1, T5, T10), at the first, third, and fifth lumbar vertebrae (L1, L3, L5) and at the pelvis (the posterior-superior iliac spine). At each location, the motions on the body surface were measured in the three orthogonal axes within the sagittal plane (i.e., the vertical, fore-and-aft, and pitch axes). The force at the seat surface was also measured. Frequency response functions (i.e., transmissibilities and apparent mass) were used to represent the responses of the body. Non-linear characteristics were observed in the apparent mass and in the transmissibilities to most measurement locations. Resonance frequencies in the frequency response functions decreased with increases in the vibration magnitude (e.g. for the vertical transmissibility to L3, a reduction from 6.25 to 4.75 Hz when the vibration magnitude increased from 0.125 to 2.0 ms(-2) r.m.s.). The transmission of vibration within the spine also showed some evidence of a non-linear characteristic. It can be concluded from this study that the dynamic responses of seated subjects are clearly non-linear with respect to vibration magnitude, whereas previous studies have reported inconsistent conclusions. More understanding of the dependence on vibration magnitude of both the dynamic responses of the soft tissues of the body and the muscle activity (voluntary and involuntary) is required to identify the causes of the non-linear characteristics observed in this study.     

Modeling the human body/seat system in a vibration environment

The vibration environment is a common man-made artificial surrounding with which humans have a limited tolerance to cope due to their body dynamics. This research studied the dynamic characteristics of a seated human body/seat system in a vibration environment. The main result is a multi degrees of freedom lumped parameter model that synthesizes two basic dynamics: (i) global human dynamics, the apparent mass phenomenon, including a systematic set of the model parameters for simulating various conditions like body posture, backrest, footrest, muscle tension, and vibration directions, and (ii) the local human dynamics, represented by the human pelvis/vibrating seat contact, using a cushioning interface. The model and its selected parameters successfully described the main effects of the apparent mass phenomenon compared to experimental data documented in the literature. The model provided an analytical tool for human body dynamics research. It also enabled a primary tool for seat and cushioning design. The model was further used to develop design guidelines for a composite cushion using the principle of quasi-uniform body/seat contact force distribution. In terms of evenly distributing the contact forces, the best result for the different materials and cushion geometries simulated in the current study was achieved using a two layer shaped geometry cushion built from three materials. Combining the geometry and the mechanical characteristics of a structure under large deformation into a lumped parameter model enables successful analysis of the human/seat interface system and provides practical results for body protection in dynamic environment.     

Dynamic forces over the interface between a seated human body and a rigid seat during vertical whole-body vibration

Biodynamic responses of the seated human body are usually measured and modelled assuming a single point of vibration excitation. With vertical vibration excitation, this study investigated how forces are distributed over the body-seat interface. Vertical and fore-and-aft forces were measured beneath the ischial tuberosities, middle thighs, and front thighs of 14 subjects sitting on a rigid flat seat in three postures with different thigh contact while exposed to random vertical vibration at three magnitudes. Measures of apparent mass were calculated from transfer functions between the vertical acceleration of the seat and the vertical or fore-and-aft forces measured at the three locations, and the sum of these forces. When sitting normally or sitting with a high footrest, vertical forces at the ischial tuberosities dominated the vertical apparent mass. With feet unsupported to give increased thigh contact, vertical forces at the front thighs were dominant around 8 Hz. Around 3–7 Hz, fore-and-aft forces at the middle thighs dominated the fore-and-aft cross-axis apparent mass. Around 8–10 Hz, fore-and-aft forces were dominant at the ischial tuberosities with feet supported but at the front thighs with feet unsupported. All apparent masses were nonlinear: as the vibration magnitude increased the resonance frequencies decreased. With feet unsupported, the nonlinearity in the apparent mass was greater at the front thighs than at the ischial tuberosities. It is concluded that when the thighs are supported on a seat it is not appropriate to assume the body has a single point of vibration excitation.     

Measurement and modelling of x-direction apparent mass of the seated human body-cushioned seat system

For modelling purposes and for evaluation of driver's seat performance in the vertical direction various mechano-mathematical models of the seated human body have been developed and standardized by the ISO. No such models exist hitherto for human body sitting in an upright position in a cushioned seat upper part, used in industrial environment, where the fore-and-aft vibrations play an important role. The interaction with the steering wheel has to be taken into consideration, as well as, the position of the human body upper torso with respect to the cushioned seat back as observed in real driving conditions. This complex problem has to be simplified first to arrive at manageable simpler models, which still reflect the main problem features. In a laboratory study accelerations and forces in x-direction were measured at the seat base during whole-body vibration in the fore-and-aft direction (random signal in the frequency range between 0.3 and 30 Hz, vibration magnitudes 0.28, 0.96, and 2.03 ms(-2) unweighted rms). Thirteen male subjects with body masses between 62.2 and 103.6 kg were chosen for the tests. They sat on a cushioned driver seat with hands on a support and backrest contact in the lumbar region only. Based on these laboratory measurements a linear model of the system-seated human body and cushioned seat in the fore-and-aft direction has been developed. The model accounts for the reaction from the steering wheel. Model parameters have been identified for each subject-measured apparent mass values (modulus and phase). The developed model structure and the averaged parameters can be used for further bio-dynamical research in this field.     

The tactile-stimulated startle response of tadpoles: acceleration performance and its relationship to the anatomy of wood frog (Rana sylvatica), bullfrog (Rana catesbeiana), and American toad (Bufo americanus) tadpoles

I described the tactile-stimulated startle response (TSR) of wood frog (Rana sylvatica), bullfrog (Rana catesbeiana), and American toad (Bufo americanus) tadpoles. One purpose was to rank species in terms of maximum acceleration performance. Also, I tested whether anatomical indicators of performance potential were predictive of realized performance. TSRs were elicited in a laboratory setting, filmed at 250 Hz, and digitally analyzed. TSRs began with two, initial body curls during which tadpoles showed a broad spectrum of movement patterns. TSR performance was quantified by maximum linear acceleration and maximum rotational acceleration of the head/body, both of which tended to occur immediately upon initiation of motion (< 0.012 sec into the response). Bullfrog tadpoles had higher maximum acceleration than the other species, but other interspecific differences were not significant. The species' rank order for the anatomical indicator of linear acceleration potential was bullfrog > wood frog > American toad. The species' rank order for the anatomical indicator of rotational acceleration potential was bullfrog > wood frog = American toad. Thus, the anatomical indicators roughly predicted the rank order of interspecific average performance. However, the anatomical indicators did not correlate with individual tadpole performance. Variability in behavioral patterns may obscure the connection between anatomy and performance. This is seen in the current lack of intraspecific correlation between a morphological indicator of acceleration capacity and acceleration performance.     

Seated whole body vibrations with high-magnitude accelerations--relative roles of inertia and muscle forces

Reliable computation of spinal loads and trunk stability under whole body vibrations with high acceleration contents requires accurate estimation of trunk muscle activities that are often overlooked in existing biodynamic models. A finite element model of the spine that accounts for nonlinear load- and direction-dependent properties of lumbar segments, complex geometry and musculature of the spine, and dynamic characteristics of the trunk was used in our iterative kinematics-driven approach to predict trunk biodynamics in measured vehicle's seat vibrations with shock contents of about 4g (g: gravity acceleration of 9.8m/s(2)) at frequencies of about 4 and 20Hz. Muscle forces, spinal loads and trunk stability were evaluated for two lumbar postures (erect and flexed) with and without coactivity in abdominal muscles. Estimated peak spinal loads were substantially larger under 4Hz excitation frequency as compared to 20Hz with the contribution of muscle forces exceeding that of inertial forces. Flattening of the lumbar lordosis from an erect to a flexed posture and antagonistic coactivity in abdominal muscles, both noticeably increased forces on the spine while substantially improving trunk stability. Our predictions clearly demonstrated the significant role of muscles in trunk biodynamics and associated risk of back injuries. High-magnitude accelerations in seat vibration, especially at near-resonant frequency, expose the vertebral column to large forces and high risk of injury by significantly increasing muscle activities in response to equilibrium and stability demands.     

Head tilt–translation combinations distinguished at the level of neurons

Angular and linear accelerations of the head occur throughout everyday life, whether from external forces such as in a vehicle or from volitional head movements. The relative timing of the angular and linear components of motion differs depending on the movement. The inner ear detects the angular and linear components with its semicircular canals and otolith organs, respectively, and secondary neurons in the vestibular nuclei receive input from these vestibular organs. Many secondary neurons receive both angular and linear input. Linear information alone does not distinguish between translational linear acceleration and angular tilt, with its gravity-induced change in the linear acceleration vector. Instead, motions are thought to be distinguished by use of both angular and linear information. However, for combined motions, composed of angular tilt and linear translation, the infinite range of possible relative timing of the angular and linear components gives an infinite set of motions among which to distinguish the various types of movement. The present research focuses on motions consisting of angular tilt and horizontal translation, both sinusoidal, where the relative timing, i.e. phase, of the tilt and translation can take any value in the range −180° to 180°. The results show how hypothetical neurons receiving convergent input can distinguish tilt from translation, and that each of these neurons has a preferred combined motion, to which the neuron responds maximally. Also shown are the values of angular and linear response amplitudes and phases that can cause a neuron to be tilt-only or translation-only. Such neurons turn out to be sufficient for distinguishing between combined motions, with all of the possible relative angular–linear phases. Combinations of other neurons, as well, are shown to distinguish motions. Relative response phases and in-phase firing-rate modulation are the key to identifying specific motions from within this infinite set of combined motions.     

Environmental stressors during space flight: potential effects on body temperature

1. Organisms may be affected by many environmental factors during space flight, e.g., acceleration, weightlessness, decreased pressure, changes in oxygen tension, radiofrequency radiation and vibration. 2. Previous studies of change in body temperature--one response to these environmental factors--are reviewed. 3. Conditions leading to heat stress and hypothermia are discussed.     

The responses of auditory ventral-cord neurons ofLocusta migratoria to vibration stimuli

Summary Most of the auditory neurons in the ventral nerve cord ofLocusta migratoria carry information not only from the tympanal organs but also from the subgenual organs (vibration sensors). Six of the eight neuron types studied electrophysiologically respond to at least these two modalities. Artificial sounds (white noise and pure tones varying in frequency and intensity) and sinusoidal vibration (200 Hz with an acceleration of 15.8 cm/s₂ or 2000 Hz and 87 cm/s₂) were used as stimuli.Complex excitatory and/or inhibitory interactions of the signals from both tympanal organs form the discharge patterns of auditory ventral-cord neurons in response to stimulation with air-borne sound. Normally the input of the ipsilateral sense organ dominates. The response patterns of these same neurons elicited by vibration stimuli are formed differently, as follows: (1) the sensory inputs of all subgenual organs are integrated in the responses of the ventral-cord neurons; in a single neuron they have either excitatory or inhibitory effects, but not both. (2) The more legs vibrated, the larger is the response. (3) The subgenual organs in the middle legs are most effective, those in the hind legs least so. (4) Ipsilateral vibration has more effect than contralateral.The six auditory neurons react to vibration combined with air-borne sound in different ways. The B neuron is the only one inhibited by vibration stimuli. The G neuron has been studied more intensively; because its anatomical arrangement and the location of the endings of the subgenual receptor fibers are known, it could be inferred from effects of transection of the connectives that interneurons are interposed between receptor cells and the G neuron.Part of the program Sonderforschungsbereich 114 (Bionach) Bochum, under the auspices of the Deutsche Forschungsgemeinschaft, with the support of the Slovenic Research Society (RSS)     

Trunk muscle electromyography and whole body vibration

By measuring the electromyographic (EMG) activity of the paraspinal muscles, we have estimated the average and peak-to-peak torque imposed on the spine during whole body vibration. Six subjects had surface electrodes placed on their erector spinae muscles at the L3 level. The EMG-torque relationship was estimated by having each subject perform isometric horizontal pulls in an upright seated posture. The subject was then vibrated vertically and sinusoidally in a controlled, flexed, slightly lordotic seated posture, in 1 Hz increments from 3 to 10 Hz at a 0.1 g RMS seat acceleration level. Between vibration readings taken at each frequency, a static reading was also taken with the subject maintaining the same posture. The entire vibration-static 3-10 Hz test was repeated for reliability purposes. Specialized digital signal processing techniques were developed for the EMG signals to enhance the measured cyclic muscle activity and to allow automatic measurement of the time relationship between the mechanical displacement and the estimated torque. We found significantly more average and peak-to-peak estimated torque at almost all frequencies for vibration vs static sitting.     

Effects of vibration on arm and shoulder muscles in three body postures

The electromyographic responses of arm and shoulder muscles to vibrations were studied in three postures similar to the postures of drilling in a ceiling, drilling in a wall and drilling in a floor. This experiment was performed within the defined parameters of: vibrational frequency at 30 Hz, acceleration level 40 m.s-2 (rms), pushing force expressed as percentage maximal voluntary contraction, and gripping force which was set at 100 N. The exposure time for each test was 5 min. The general findings from these three body postures show that all the examined muscles were affected by exposure to vibration. The EMG index increased as follows: trapezius muscle 39% (p less than 0.05), lower-arm flexor muscles 23% (p less than 0.05), infraspinatus muscle 14% (p less than 0.05), lower-arm extensor muscles 14% (p less than 0.1) and biceps muscle 6% (p less than 0.1). The muscle most affected by vibration was found to be the trapezius muscle. It should be taken into consideration that vibration can be a contributing factor in neck/shoulder disorders among power handtool operators. The general conclusion from this study is that changes in working posture give different transmissions of vibration in the upper extremities. It seems as if the prime movers and muscles with an increased muscle length or increased degree of contraction are most affected by vibration.     

Spatio-temporal convergence (STC) in otolith neurons

It has been recently demonstrated that some primary otolith afferents and most otolith-related vestibular nuclei neurons encode two spatial dimensions that can be described by two vectors in temporal and spatial quadrature. These cells are called broadly-tuned neurons. They are characterized by a non-zero tuning ratio which is defined as the ratio of the minimum over the maximum sensitivity of the neuron. Broadly-tuned neurons exhibit response gains that do not vary according to the cosine of the angle between the stimulus direction and the cell's maximum sensitivity vector and response phase values that depend on stimulus orientation. These responses were observed during stimulation with pure linear acceleration and can be explained by spatio-temporal convergence (STC) of primary otolith afferents and/or otolith hair cells. Simulations of STC of the inputs to primary otolith afferents and vestibular nuclei neurons have revealed interesting characteristics: First, in the case of two narrowly-tuned input signals, the largest tuning ratio is achieved when the input signals are of equal gain. The smaller the phase difference between the input vectors, the larger the orientation differences that are required to produce a certain tuning ratio. Orientation and temporal phase differences of 30–40° create tuning ratios of approximately 0.10–0.15 in target neurons. Second, in the case of multiple input signals, the larger the number of converging inputs, the smaller the tuning ratio of the target neuron. The tuning ratio depends on the number of input units, as long as there are not more than about 10. For more than 10–20 input vectors, the tuning ratio becomes almost independent of the number of inputs. Further, if the inputs comprise two populations (with different gain and phase values at a given stimulus frequency), the largest tuning ratio is obtained when the larger population has a smaller gain. These findings are discussed in the context of known anatomical and physiological characteristics of innervation patterns of primary otolith afferents and their possible convergence onto vestibular nuclei neurons.     

Study of Biomechanical Response of Human Hand-Arm to Random Vibrations of Steering Wheel of Tractor

This paper reports a study on the biomechanical response of a human hand-arm model to random vibrations of the steering wheel of a tractor. An anatomically accurate bone-only hand-arm model from TurboSquidTM was used to obtain a finite element (FE) model to understand the Hand-arm vibration syndrome (HAVS), which is a neurological and vascular disorder caused by exposure of the human hand-arm to prolonged vibrations. Modal analysis has been done to find out the first few natural frequencies and mode shapes of the system. Coupling of degrees of freedom (DOF) had to be done in the FE idealization to do modal analysis, as the bones were not attached to each other in the TurboSquidTM model. The shoulder bone, scapula, has been constrained at one end for eigenvalue analysis. It was observed that the first five natural frequencies were in the range of 0-250 Hz, which is the range in which the effect of HAVS is the highest. Harmonic analysis was done by giving a swept sine excitation in the frequency range 0 to 200 Hz. For this, a force input of 25 N was imparted at nodes perpendicular to the hand, the force value chosen being the nominal force in most applications involving powered hand-held tools and steering wheels of tractors. The nodes chosen for force application were determined experimentally from observations made by gripping the steering wheel. The frequency response function (FRF) plots were obtained in the x, y and z directions. Random vibration analysis was done next by giving force power spectral densities (PSD) in the form of nodal excitation as input to the FE model of hand-arm, and computing the output acceleration PSDs. The input force PSDs were measured using FlexiForce® sensors along the three axes. The acceleration responses at the steering wheel were also measured using tri-axial accelerometers for validating the computed results. The output acceleration PSDs were then weighted using the frequency weighting curves for hand-arm vibration and the total daily exposure A(8), computed using ISO 5349-1 standards, was compared with the vibration action and limit values. The A(8) values obtained are found to be higher than the vibration limit values.     

Reliability of assessing postural control during seated balancing using a physical human-robot interaction

This study evaluated the within- and between-visit reliability of a seated balance test for quantifying trunk motor control using input–output data. Thirty healthy subjects performed a seated balance test under three conditions: eyes open (EO), eyes closed (EC), and eyes closed with vibration to the lumbar muscles (VIB). Each subject performed three trials of each condition on three different visits. The seated balance test utilized a torque-controlled robotic seat, which together with a sitting subject resulted in a physical human-robot interaction (pHRI) (two degrees-of-freedom with upper and lower body rotations). Subjects balanced the pHRI by controlling trunk rotation in response to pseudorandom torque perturbations applied to the seat in the coronal plane. Performance error was expressed as the root mean square (RMSE) of deviations from the upright position in the time domain and as the mean bandpass signal energy (E_mb) in the frequency domain. Intra-class correlation coefficients (ICC) quantified the between-visit reliability of both RMSE and E_mb. The empirical transfer function estimates (ETFE) from the perturbation input to each of the two rotational outputs were calculated. Coefficients of multiple correlation (CMC) quantified the within- and between-visit reliability of the averaged ETFE. ICCs of RMSE and E_mb for all conditions were ≥0.84. The mean within- and between-visit CMCs were all ≥0.96 for the lower body rotation and ≥0.89 for the upper body rotation. Therefore, our seated balance test consisting of pHRI to assess coronal plane trunk motor control is reliable.     

The transmission of translational seat vibration to the head--II. Horizontal seat vibration

The second part of this study of the six axes of head motion caused by translational seat vibration is concerned with the effect of fore-and-aft (x-axis) and lateral (y-axis) seat vibration. Seat-to-head transmissibilities have been determined at frequencies up to 16 Hz for each of the three translational and three rotational axes of the head during exposure to random vibration of the seat. Repeatability measures within a single subject and studies of the variability across a group of twelve subjects have been conducted with two seating conditions: a rigid seat with a backrest, and the same seat with no backrest. Fore-and-aft seat motion mainly resulted in head motion within the mid-sagittal plane (x-z plane). Without the backrest, transmissibilities for the fore-and-aft, vertical and pitch axes of the head were greatest at about 2 Hz. The backrest greatly increased head vibration at frequencies above 4 Hz and caused a second peak in the transmissibility curves at about 6 to 8 Hz. Lateral seat motion mainly caused lateral head motion with a maximum transmissibility at about 2 Hz. The backrest had little effect on the transmission of lateral vibration to the head. For both axes of excitation inter-subject variability was much greater than intra-subject variability.     

Criteria of quality and methods of measurement in multichannel biotelemetry.

In contrast with one-channel biotelemetry, multichannel biotelemetry systems permit the analysis of biological systems by means of cybernetics. Some simple examples are given. The basic concept of a multichannel system is described. The criteria of quality of which the nonevident ones are defined by detailing the test circuits are the following (input criteria): number of channels, bandwidth, voltage range, impedance, maximum voltage outside the signal band, stability of the power supply for transducers, factor of safety against multipath propagation, distance between transmitter and receiver, shock, vibration and temperature range, weight, volume and operating time. Output criteria: signal-to-noise ratio, linearity, crosstalk transfer function, different time lags.     

Repeatable measures of take-off flight performance in auklets

Rapid acceleration is the key to a successful escape manoeuvre and has attracted considerable research attention in a wide array of taxa. I recorded take-offs of least auklets Aethia pusilla and crested auklets Aethia cristatella with digital video (60 frames per second). To smooth time–location data derived from video, I used predicted mean square error quintic splines, which have been shown to be good predictors of true acceleration. Repeated recordings of the same individual bird allowed me to measure repeatability of take-off acceleration and velocity to find the most robust and biologically meaningful measure. The most repeatable take-off parameters were power at time t =0.17 s after take-off ( r =75%) and acceleration at t =0.17 s ( r =72%). The horizontal component of velocity at t =0.32 s was least affected by the slope of the take-off trajectory. The mean acceleration of both species is close to expected values based on body mass, even though all previously studied species had considerably lower body mass. Within least auklets, however, I did not find a significant relationship of velocity or acceleration with mass. This would be expected if the observed drop in mass after hatching was an adaptation to reduce the risk of predation. I conclude that acceleration and exerted power at a certain time after take-off is repeatable and the most suitable measure of performance for both inter- and intra-specific comparisons.     

Can Ethograms Be Automatically Generated Using Body Acceleration Data from Free-Ranging Birds?

An ethogram is a catalogue of discrete behaviors typically employed by a species. Traditionally animal behavior has been recorded by observing study individuals directly. However, this approach is difficult, often impossible, in the case of behaviors which occur in remote areas and/or at great depth or altitude. The recent development of increasingly sophisticated, animal-borne data loggers, has started to overcome this problem. Accelerometers are particularly useful in this respect because they can record the dynamic motion of a body in e.g. flight, walking, or swimming. However, classifying behavior using body acceleration characteristics typically requires prior knowledge of the behavior of free-ranging animals. Here, we demonstrate an automated procedure to categorize behavior from body acceleration, together with the release of a user-friendly computer application, “Ethographer”. We evaluated its performance using longitudinal acceleration data collected from a foot-propelled diving seabird, the European shag, Phalacrocorax aristotelis. The time series data were converted into a spectrum by continuous wavelet transformation. Then, each second of the spectrum was categorized into one of 20 behavior groups by unsupervised cluster analysis, using k-means methods. The typical behaviors extracted were characterized by the periodicities of body acceleration. Each categorized behavior was assumed to correspond to when the bird was on land, in flight, on the sea surface, diving and so on. The behaviors classified by the procedures accorded well with those independently defined from depth profiles. Because our approach is performed by unsupervised computation of the data, it has the potential to detect previously unknown types of behavior and unknown sequences of some behaviors.