Introduction

Shock level monitoring for decision-making and response initiation is critical in various applications within the health1,2,3, safety4,5, manufacturing6,7, and defense industries8,9. Among these applications, the prediction of mild Traumatic Brain Injuries (mTBIs) stands out considering the lack of feasibility in the current technology despite the importance of impact monitoring for proper prediction of a mTBI10. mTBIs or concussions are brain injuries that typically occur after sustaining a head trauma and can lead to serious health complications11. Therefore, early diagnosis of concussions is critical to provide immediate care for urgent cases before complications can develop12. Many of the wearable technologies used for concussion diagnosis rely on accelerometers, which require continuous power for data collection and signal processing13,14. These systems typically drain power continuously, which reduces their portability and suitability for this application. Inertial switches provide more compact, lightweight, and energy-efficient alternatives to such systems, as they only consume power once they are triggered and can eliminate the need for signal processing and other operations required in accelerometer-based systems15.

Inertial switches can be classified based on their sensing direction as unidirectional16,17,18; bidirectional19,20,21,22; quad-directional23,24,25; or omnidirectional26,27,28,29,30,31. One important consideration for omnidirectional devices is the uniformity of the spring stiffness across their circumference to ensure consistent acceleration thresholds across all angles of the sensing plane, regardless of the direction of the applied accelerations32. This uniformity of acceleration detection is especially important in applications such as the early prediction of mTBIs because, in cases such as collisions in American football, the impacts can come from any direction and have a wide range of magnitudes. Among the omnidirectional inertial switches presented in the literature with experimental results to compare thresholds across different planar angles, the best uniformity values achieved were reported by Du et al.27 and Chen et al.33, who reported maximum variation in thresholds across 15° angle rotational steps of 3.39% and 3.45%, respectively.

Recently, several inertial switches have been presented in the literature with designs that allow them to detect the triggering of multiple thresholds rather than a single threshold34,35,36,37. This triggering capability makes it possible to configure the switch to perform different actions for each threshold trigger, enabling autonomous responses based on the severity of impact20. This feature can also provide a rough indication of the level of impact rather than merely alerting if a specific acceleration threshold was exceeded. This quantitative detection can be especially beneficial in the prediction of concussions as it can provide an estimate of range of the trauma that was sustained, considering that no specific threshold exists beyond which an injury is guaranteed to occur. Instead, brain injury is affected by several factors, such as the number of collisions, genetics, and age38,39,40,41. Thus, Zhang et al.42 conducted a study to predict the likelihood of a mTBI at different shock magnitudes and estimated probabilities of 25%, 50%, and 80% at acceleration magnitudes of 50 ɡ, 80 ɡ, and 100 ɡ, respectively. Utilization of multi-threshold inertial switches that operate within similar ranges can provide critical support for the prediction of mTBIs after collisions to initiate appropriate responses immediately when required.

The only three multi-threshold inertial switches from the literature that offer sensitivity in more than two directions were proposed by our research team for high-ɡ applications23,24 and low-ɡ applications25. While these designs enable shock detection in four directions, they do not achieve true omnidirectional sensitivity. In contrast, advancements by Du et al.27 and Chen et al.33 demonstrated high uniformity in omnidirectional sensitivity, yet their devices are limited to a single threshold. The introduction of a second threshold poses a challenge for planar inertial switches when high uniformity is desired due to the overall increase in system stiffness after contact is made between the proof mass and the stationary electrode/s of the first acceleration threshold. Consequently, the literature lacks a truly in-plane omnidirectional inertial switch with multiple thresholds.

In this work, a 2D omnidirectional monolithic MEMS inertial switch with multiple thresholds for assistance with early prediction of concussions is presented (see Fig. 1). The design was modeled using SOLIDWORKS Simulation software to optimize the dimensions and geometries to achieve high uniformity and the desired threshold values. The device was fabricated through the commercially available SOIMUMPs process, followed by additional gold sputtering performed in-house to increase contact conductivity. Testing was performed using a drop-table system and the results demonstrate the device has the capacity to consistently detect two distinct acceleration thresholds in response to mechanical shocks from any direction within the sensing plane. The experimental results showed high uniformity across all angles with the first and second threshold values aligning with the shock magnitude ranges related to the mTBIs mentioned previously.

Fig. 1
figure 1

Application of the inertial switch in sports to immediately detect possible brain injuries.

Full size image

Design and fabrication

The inertial switch was designed by our team using computer-aided design (CAD) tools and was primarily refined through finite element analysis (FEA) with SOLIDWORKS Simulation software. We utilized this software to identify the optimal dimensions and simulate ideal performance to meet the design goals. In a FEA, a 3D model incorporating linear elastic material properties is used to accurately simulate structural stress and optimize geometrical parameters, such as width, length, thickness, and angles, to achieve the best values. However, these optimization efforts are limited by the fabrication constraints of the Silicon On Insulator Multi-User MEMS Processes (SOIMUMPs) from Science, Inc., the foundry of the MEMS fabrication43.

Design process

The goal of this switch design was to detect axial and off-axial (oblique) accelerations from all directions within the 2D sensing plane. Therefore, the design process primarily focused on modeling and optimizing the suspended springs to achieve consistent stiffness in all directions within this plane. This consistency was important to guarantee uniform responses to shocks coming from any angle. Three different spring designs were evaluated to find the optimal uniform stiffness profile distribution of the mechanical stiffness under a 100 µN load across 360° (see Fig. 2). A serpentine spring design from the literature44 and our initial asymmetrical semi-circle spring design, represented as Cases 1 and 2, exhibited non-uniform stiffness results across the 2D plane. In contrast, our S-shaped spring design (Case 3) achieved uniform stiffness results in all directions. Table 1 lists the properties of the silicon material used for the simulation, as specified in the foundry’s fabrication guide43.

Fig. 2
figure 2

Simulated comparison of three different spring design cases of the inertial switch (top) and their stiffness distribution profiles under a 100 µN load across 360° (bottom).

Full size image
Table 1 Silicon material properties used for the simulation.
Full size table

Finalized design

After fabricating three different design versions that failed during testing (see Fig.S1 in the Supplementary Material), the fourth version was identified as the optimal design, illustrated in Fig. 3. This final design features a proof mass suspended by four S-shaped springs arranged symmetrically at 90° angles (refer to Case 3 in Fig. 2). These quad-springs, formed by two semi-circles joined in an S-shape, enhance the detection uniformity for both axial and off-axial accelerations within the sensing plane. The design includes a proof mass with four circular edges, surrounded by a set of movable circular edges that represent the first acceleration threshold, followed by an additional similar set of fixed circular edges for the second threshold, aligned concentrically to the first switch electrodes. This circular contact configuration between the mass and first and second triggers enhances the uniformity of detection for axial or off-axial (oblique) accelerations from all in-plane directions. This design concept can also be expanded to include additional thresholds for various applications. The dimensions and geometrical parameters of the inertial switch are given in Table 2.

A key aspect of the design is the use of a movable S-shaped spring with the first threshold electrode (see Fig. 3). This optimized spring design has low stiffness, which minimizes the new added stiffness to the overall stiffness of the device. By reducing this accumulated stiffness, the design maintains uniform stiffness in all 2D directions beyond the first threshold.

When the inertial switch experiences axial or oblique shocks surpassing its set acceleration levels, the proof mass rapidly moves and makes contact with the stationary electrodes (Switches 1 and 2). Initially, this contact occurs at the first acceleration threshold, forming an electrical path and generating a trigger signal. The trigger signal for Switch 1 is activated based on an OR logic condition when the proof mass makes contact with any of the four first stationary electrodes (S1X±, S1Y±) (see Fig. 4). The activation process for Switch 2 relies on the pre-activation of Switch 1 with an AND logic condition to the following condition. The following condition is an OR logic condition involving the activation of any of the second four fixed electrodes (S2X±, S2Y±) (see Fig. 4).

Fig. 3
figure 3

2D omnidirectional inertial switch with multiple thresholds: (a) structural sketch and close-up views (b and c) of the of the first and second switches from different angles.

Full size image
Fig. 4
figure 4

Schematic diagram illustrating the electrical working principle for activation of Switches 1 and 2. The upper table on the right shows the OR gate logic for (S1X±, S1Y±) to activate Switch 1. The bottom table shows the AND gate logic for Switch 1, followed by the OR gate logic for (S2X±, S2Y±) to activate Switch 2.

Full size image
Table 2 Geometric dimensions of the inertial switch design.
Full size table

Basic physical model and theoretical analysis

A simplified physical model of the designed switch is illustrated in Fig. 5. Initially, before any acceleration is applied, the proof mass and first threshold springs are in their original lengths without elongation (Fig. 5). As a shock acceleration is applied, the proof mass begins to move toward the stationary electrodes. If the acceleration exceeds the first threshold, then the proof mass continues its movement until it closes the first electrode gap (x1) and reaches the first threshold level (Fig. 5). At this point, the stiffness of the first threshold springs is added to the overall stiffness of the system, making the device stiffer and requiring higher acceleration for further movement. As the acceleration passes this initial threshold, the proof mass will move while it is still engaged to the first threshold springs (Fig. 5). Upon reaching the second threshold, the device will hit the second threshold level and remain there until the acceleration decreases (Fig. 5).

Fig. 5
figure 5

Simplified schematic of the physical model of the omnidirectional inertial switch.

Full size image

An analytical method is introduced to determine the timing of the switching event, which is crucial for studying the dynamic behavior of the designed inertial switch. When an acceleration is applied to the spring-mass system in any direction, the equation for the motion of the proof mass is given by:

$$\:m\ddot{x}+\left({k}_{1}+{k}_{2}v\left(x\right)\right)x=ma\left(t\right)$$
(1)
$$\:\:\:\:\:\:\:\:\:\:v\left(x\right)=\left\{\begin{array}{c}0,\:\:x<{x}_{1}\\\:1,\:\:x\ge\:{x}_{1}\end{array}\right.$$
(2)

In this equation, represents the mass of the proof mass, 1 denotes the total spring constant of the four S-springs supporting the proof mass, and 2​ is the additional spring constant at the first threshold. As the added stiffness 2 is not activated unless the proof mass touches the first threshold at the displacement of x1, then v(x) represents the activation step function as expressed in Eq. (2).

In the majority of shock and impact scenarios, the acceleration signal () can be modeled as a half-sine wave function characterized by amplitude 0 ​and pulse width 0, and is expressed as follows:

$$\:\:\:\:\:\:\:\:\:\:a\left(t\right)=\left\{\begin{array}{c}{a}_{0}\:sin\left(\frac{\pi\:\:t}{{t}_{0}}\right),\:\:0\le\:t\le\:{t}_{0}\\\:0\:,\:\:t>{t}_{0}\end{array}\right.$$
(3)

The ode45 numerical solver of MATLAB software was employed to simulate the acceleration pulse as a half-sine wave with a duration of 3 ms. To determine the acceleration thresholds, the mechanical system of the inertial switch (Eq. 1) is subjected to various axial accelerations in the x–y plane with amplitudes of 25 ɡ, 35 ɡ, 90 ɡ, and 108 ɡ (see Fig. 6). The black dashed lines indicate the activation thresholds for Switches 1 and 2. In Fig. 6, the proof mass does not make contact with the first switch electrode when an acceleration of 25 ɡ is applied. However, as illustrated in Fig. 6, at an applied acceleration of 35 ɡ, the proof mass reaches the distance x1 (3 μm) between the proof mass and first stationary electrodes in Switch 1, indicating the first threshold acceleration and resulting in the ON state of Switch 1. In Fig. 6, under 90 ɡ acceleration, the proof mass contacts and moves past the first stationary electrode, indicating ON and OFF states for Switches 1 and 2, respectively. When the applied shock is increased to 108 ɡ, as seen in Fig. 6, the simulated second acceleration threshold is reached, activating Switch 2 to the ON state.

Fig. 6
figure 6

Numerical simulation of the mass-spring system under different acceleration loads with a duration of 3 ms.

Full size image

The applied acceleration pulse can be modeled as a static or dynamic load and can be classified based on the corresponding normalized natural period value, as reported in the literature45. Therefore, the system behaves in a quasi-static excitation mode if its loading period is equal to or more than five times the natural period of the structure (see Fig. 7). Under this condition, then the static FEA can accurately represent the response of the device. Thus, it is advantageous for acceleration switches to have considerably shorter natural periods than the duration of the impacts in their targeted applications (e.g., 15–20 ms for concussions46, to ensure the same consistent acceleration thresholds. For the device presented, the modal simulation determines the natural period to be 0.56 ms.

Fig. 7
figure 7

Relative response spectrum illustrating the transient and quasi-static behavior regimes for different shock durations: (a) overall shock spectrum, (b) dynamic response for the transient regime, and (c) system response for the quasi-static regime.

Full size image

Fabrication

The device was fabricated using the commercial SOIMUMPs process by Science, Inc. (formerly MEMSCAP)43. This fabrication technology enables the integration of the proof mass, springs, and electrodes in a single plane. To enhance the conductivity of the electrodes, 10 nm and 120 nm layers of titanium (Ti) and gold (Au), respectively, were sputtered on to the top and side surfaces of the device in-house at the KAUST Nanofabrication Lab. The fabricated device consists of a silicon structural layer with anchored metal probe pads (see Fig. 8). The fabrication of the inertial switch using SOIMUMPs involved several steps on a silicon-on-insulator (SOI) wafer, as detailed in the foundry’s guidebook43. The fabrication process began with a doped SOI wafer, which included a 25 μm device layer, 2 μm silicon oxide (SiO2) interlayer, and a 400 μm handle layer. The main fabrication steps (see Fig. 9), are as follows:

  1. I.

    Metal Deposition: Using the liftoff process, a 20 nm chromium (Cr) layer and a 500 nm gold (Au) layer were deposited on the top surface of the device layer through Electron beam evaporation (E-beam evaporation), forming the electrical pads.

  2. II.

    Top-side Etching: Deep reactive ion etching (DRIE) was used to etch through the device layer and form the inertial switch components with a thickness of 25 μm.

  3. III.

    Backside Etching: DRIE was further used to etch through the bottom handle layer until the etching process reached the buried oxide layer (SiO2) at a depth of 400 μm. During this step, a protective material was applied to the front side to safeguard the device layer.

  4. IV.

    Final Release: After the front side protection material was removed (specific details not documented by the foundry), vapor-phase hydrofluoric acid was employed to selectively etch away at the suspended oxide layer (SiO2), completing the release of the inertial switch components.

Fig. 8
figure 8

Scanning Electron Microscope (SEM) micrograph of the fabricated omnidirectional inertial switch.

Full size image
Fig. 9
figure 9

Top view and cross-sectional view illustrating the main steps of the SOIMUMPs process for fabricating the inertial switch43.

Full size image

Finite element analysis (FEA)

A finite element (FE) model of the inertial switch was built using SOLIDWORKS Simulation software to achieve optimal performance and evaluate the experimental data. To accurately simulate structural stress and determine the acceleration values that meet the design requirements, a static FEA of the 3D model of the device was performed using SOLIDWORKS Simulation software. The simulation used a free tetrahedral mesh with a maximum finer size and the default solver settings. The simulation included mechanical fixed boundary conditions, marked in yellow in Fig. 3, and a series of contact pairs between the circular edges of the proof mass and the first movable electrodes (Switch 1) and between the first and second stationary electrodes of Switches 1 and 2, respectively. These contact pairs were applied exclusively to the lateral surfaces of the movable and stationary electrodes.

The FEA indicated a consistent first acceleration threshold of 35 ɡ in all directions, as illustrated in Fig. 10 for the two sample directions of (Y-) at θ = 0° and θ = 45°, thanks to the device’s uniform omnidirectional stiffness (refer to Case 3 in Fig. 2). For the activation of Switch 2, minor variation existed between the axial and off-axial acceleration thresholds, caused by the added stiffness resulting from the contact between Switch 1 and the proof mass, which increased the overall stiffness of the device. By strictly adhering to the fabrication rules of SOIMUMPs process, this variation was minimized to this level with the associated geometric parameters. The maximum difference in the second acceleration threshold between the symmetrical axial directions (X+, X-, Y+, Y-) and their associated 45° directions was only 3 ɡ. Switch 2 was activated at 111 ɡ and 108 ɡ in the (Y-) direction at angles of 0° and 60°, respectively (see Fig. 11). Comprehensively, the FEA results for both switches were assessed by applying acceleration at 15° increments around the full 360° range, and consistent responses were observed in both axial and off-axial (oblique) directions (see Fig. 12).

Fig. 10
figure 10

The FEA results using SOLIDWORKS Simulation demonstrating the first acceleration threshold of 35 ɡ in: (a) the (Y-) direction at θ = 0° and (b) (45°) direction.

Full size image
Fig. 11
figure 11

The FEA results using SOLIDWORKS Simulation showing the second acceleration threshold in the (0°) and (60°) directions, respectively: (a) 111 ɡ and (b) 108 ɡ.

Full size image
Fig. 12
figure 12

The FEA results using SOLIDWORKS Simulation for the first and second acceleration thresholds, evaluated across 360° at 15° increments.

Full size image

Testing setup and results

The performance of the fabricated inertial switch device was thoroughly characterized and tested using several laboratory tools tailored to specific testing objectives. Initially, we assessed the response of the device to multiple shock levels resulting in the identification of distinct acceleration thresholds of the device across all directions. Subsequently, we investigated the device’s dynamic behavior by identifying the primary resonant frequencies along with their corresponding vibrational mode shapes.

Shock testing setup

The fabricated device was wire-bonded to a PCB and tested using a Lansmont drop-table system. The experimental setup and schematic configuration are shown in Fig. 13. Calibration and acceleration measurements of the tested device were performed using a reference accelerometer and trigger signals from the inertial switch were captured with an Agilent 6000 MSO6034A oscilloscope. Figure 13 shows the shock test setup, with a close-up view of the tested device fixed on a 3D-printed fixture mounted on the side of the drop-table. Figure 13 illustrates the connection between the power source, tested switch, and multi-channel oscilloscope through a schematic diagram of the test circuit.

For testing, the first threshold switch was activated by the activation of any trigger of the four first triggers (S1X±, S1Y±) using an OR logic condition (see Fig. 13). This configuration allowed the detection of the axial and off-axial accelerations uniformly, covering all of the omnidirectional inputs within the sensing plane. Once Switch 1 was activated, Switch 2 was similarly activated based on the OR logic condition for any of the four second triggers (S2X±, S2Y±), ensuring comprehensive omnidirectional sensing in the sensitive plane.

Fig. 13
figure 13

(a) The shock test setup, featuring the drop-table system and a close-up view of the device mounted on the rotating disk. (b) The schematic configuration of the test circuit. (c) The circuit configuration for the first threshold level switch, which mirrors the setup for the second threshold level switch.

Full size image

To induce a shock, the drop-table was raised to a pre-set height using a hoist and then released, allowing drop-table to accelerate freely until it collided with the hydraulic base at the bottom. This collision produced a shock on the attached tested device. The magnitude of the shock increased with the height from which the table was dropped. The duration of the shock was influenced by the stiffness of the cushioning sponge placed between the drop-table and the hydraulic base. The duration of the shock could be extended by adding more cushioning sheets. Typically, the drop-table system generated a half-sine shock pulse lasting from 0.5 ms to 4 ms. In this experiment, when the drop-table was freely released from a pre-set height of 6.99 cm, the tested inertial switch device experienced a shock amplitude of 95 ɡ. This pulse activated Switches 1 and 2 at an angle of θ = 60° (see Fig. 14).

Fig. 14
figure 14

Example of a shock pulse (a) generated by the drop-table system, showing an amplitude of 95 ɡ and (b) the synchronized real-time switch activation signal detected by the oscilloscope at an angle of θ = 60°, indicating the first (yellow) and second (green) acceleration thresholds of the tested device.

Full size image

Shock test results

Threshold acceleration tests were performed sequentially and repeated at 15° increments around the full 360° for both the first and second acceleration thresholds. To identify the first threshold, the pre-set height of the drop-table was increased gradually until Switch 1 was triggered at an applied acceleration of 32 ɡ (see Fig. 15) at an angle of 0°. Subsequent tests conducted at various angles around the full 360° showed consistent activation of Switch 1 in all directions, with a slight variation at 45° where it was activated at 30 ɡ (see Fig. 15).

As acceleration was increased further, Switch 2 was triggered at 100 ɡ at 0° while Switch 1 remained ON simultaneously (see Fig. 16). Tests for the second threshold at all 360° angles showed uniform activation of Switch 2, with a minor deviation at 45° where Switch 2 was triggered at 95 ɡ (see Fig. 16). The comprehensive testing results of the acceleration thresholds for Switches 1 and 2 across all 360° angles of the sensing plane are presented in Fig. 17. Figure 18 compares the experimental results with finite element simulations for all omnidirectional angles of the applied shock, demonstrating good agreement between the measurements and simulations.

The resulting average thresholds for the first and second acceleration levels are 30 ɡ and 100 ɡ, respectively, aligning with the ranges associated with concussive injuries as discussed in the introduction. These threshold values allow for three output ranges based on the mTBI risk categorization: a low-risk range from 0 to 30 ɡ with no signal is transduced, a medium-risk range from 30 to 100 ɡ, and a high-risk range for accelerations exceeding 100 ɡ.

Fig. 15
figure 15

The experimental results of the fabricated device for the first threshold acceleration were as follows: (a) 32 ɡ at an angle of θ = 0°, with the oscilloscope triggering signal shown in (b), and (c) 30 ɡ with at an angle of θ = 45°, with the corresponding oscilloscope triggering signal illustrated in (d).

Full size image
Fig. 16
figure 16

The experimental results of the fabricated device for the second threshold acceleration were as follows: (a) 100 ɡ at an angle of θ = 0°, with the oscilloscope triggering signal shown in (b), and (c) 95 ɡ at an angle of θ = 45°, with the corresponding oscilloscope triggering signal illustrated in (d).

Full size image
Fig. 17
figure 17

The experimental results of the device switching thresholds measured across 360° at 15° increments. The insets provide examples of the rotated locations of the tested device at specific angles: 45°, 90°, 135°, 180°, 225°, 270°, and 315°.

Full size image
Fig. 18
figure 18

Comparison of the FEA and experimental results using SOLIDWORKS Simulation for the first and second acceleration thresholds, evaluated across 360° at 15° increments.

Full size image

Modal testing

To understand the dynamic behavior of the device, a camera equipped with a stroboscopic video microscopy (Polytec MSA-500, Germany) was utilized to detect the device motion, especially in-plane motion47. The purpose of this vibrational test was to identify the resonant frequencies and validate that the device operates within its intended parameters, ensuring that shock testing falls within the quasi-static regime. This system measured the absolute in-plane motion of the proof mass when the device was excited by a 6.7 V alternating current (AC) in air. By sweeping the input AC frequency and simultaneously recording the maximum displacement of the proof mass, the resonant frequencies were captured. The experimental setup is depicted in Fig. 19. The testing identified the first resonant frequency of the device to be 912 Hz, a deviation of nearly 6% from the simulated frequency of 977 Hz, as plotted in Fig. 20. In addition, the testing revealed that the axial in-plane mode shape was measured at a frequency of 1784 Hz, an approximately 4% deviation from the simulated result of 1718.6 Hz. The experimental results closely align with the simulations, with minor discrepancies that were likely due to fabrication errors causing geometric irregularities and induced stresses.

The natural periods corresponding to these frequencies are approximately 1.096 ms for 912 Hz and 0.561 ms for 1784 Hz. Considering that the loading period for concussive impacts ranges from 15 to 20 ms 46, which is significantly longer than 5.48 ms (five times 1.096 ms) and 2.805 ms (five times 0.561 ms), it is evident that the device operates within the quasi-static range. This performance is consistent with the simulated device behavior depicted in Fig. 7.

Fig. 19
figure 19

The experimental setup for vibration testing.

Full size image
Fig. 20
figure 20

Comparison of the experimental and FEA results for the out-of-plane and in-plane frequencies of the inertial switch. The upper insets illustrate the simulated frequencies (997 Hz on the left and 1718 Hz on the right) along with their respective mode shapes. The lower inset displays the first animated mode shape obtained from the experimental testing.

Full size image

Summary and conclusions

This paper introduces a novel 2D omnidirectional inertial switch capable of detecting multiple acceleration thresholds uniformly from any direction within the sensing plane, and thus surpasses the capabilities of existing single-threshold omnidirectional switches. The design of the device, which was optimized through FEA, ensures uniform stiffness and accurate threshold detection across all axial and off-axial (oblique) directions in the sensing plane. Additionally, the device was numerically modeled as a spring-mass system to evaluate its dynamic response to varying shocks. Fabrication was performed using the SOIMUMPs process, which integrates all device components in the same plane, thereby simplifying the packaging. Drop-table testing results demonstrated high directional uniformity, with variations among 15° increments of only 2 ɡ and 5 ɡ for the first and second thresholds, respectively. The average first and second thresholds were found to be 30 ɡ and 100 ɡ, respectively, which are consistent with the shock magnitudes associated with concussion risk42. The results showed close agreement with the FEA results. Vibrational analysis was conducted experimentally to test the dynamic responses of the inertial switch and identified dominant resonant frequencies at 912 Hz (out-of-plane mode) and 1784 Hz (in-plane mode). With natural periods of 1.096 ms and 0.561 ms, the device operates within the quasi-static range for concussive impacts of 15–20 ms, also matching closely the simulation results. This device represents a significant enhancement of omnidirectional sensitivity, making it highly effective for triggering applications, especially in critical health and safety monitoring scenarios. Future work will focus on further refining the design for greater accuracy and exploring integration into wearable systems for real-time shock monitoring.