Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

Parameter Optimization of Optical Six-Axis Force/Torque Sensor for Legged Robots

T0 review · 4 major / 4 minor · reviewed 2026-08-08 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that a photocoupler-based, contactless six-axis force/torque sensor, optimized with a Timoshenko-beam model, can measure ground reaction forces on a quadruped robot's feet with high resolution and far less drift under…

desk verdict New optimization methodology for a photocoupler six-axis F/T sensor, but the durability comparison against the commercial sensor is confounded by mismatched force ranges and needs reframing. read the letter →

arxiv 2502.07196 v1 pith:QETDFCWU submitted 2025-02-11 cs.RO

Hyun-Bin Kim , Byeong-Il Ham , Keun-Ha Choi , Kyung-Soo Kim This is my paper · ORCID
classification cs.RO
keywords six-axisforce/torquesensorphotocouplerquadrupedrobotTimoshenkobeammodelparameteroptimizationgroundreactionforceimpactrobustnessopticalsensing
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to show that a six-axis force/torque sensor for legged robots can be built from six photocouplers and one machined aluminum T-beam, avoiding the strain gauges and capacitive plates used in commercial sensors. The authors derive an analytic Timoshenko-beam model that maps forces and torques to photocoupler displacements, then optimize the beam dimensions and detector placements to balance sensitivity against crosstalk and error. They fabricate the sensor, calibrate it against a commercial reference sensor, and mount it on a quadruped robot's foot alongside a commercial capacitive sensor of the same 40 mm size. The reported result is that after about ten thousand steps on gravel, the commercial sensor drifted up to 3.93% of full scale while the proposed sensor drifted at most 0.375%, while offering higher resolution and lower cost. If this holds, it gives legged robots a small, cheap, impact-tolerant way to measure ground reaction forces.

What carries the argument

The central object is the $6\times 6$ sensitivity matrix $G$ that relates the six applied forces and torques to the six photocoupler displacements, $\Delta d = G F$. Each entry is built from Timoshenko-beam spring constants evaluated at the actual photocoupler positions, with the T-beam modeled as a symmetric three-arm elastomer whose loading-table rotation is constrained to zero. The optimization searches over the beam dimensions $l_1$, $l_2$, $b_1$, $b_2$, $h$, the loading-table radius $r$, and the horizontal photocoupler radius $r_{s2}$, minimizing the chosen objective $\mathrm{Cond}(G)\,\Vert G \Vert_2/\Vert G \Vert_*^2$ subject to stress and 40 mm envelope constraints. The selected objective simultaneously penalizes ill-conditioning and over-large singular values while rewarding overall sensitivity, and the resulting parameters fix both the elastomer geometry and the detector positions on the PCB.

What would settle it

Apply each of the six load cases to the optimized T-beam in a finite-element simulation and compare the computed displacements at the photocoupler positions with the analytic spring constants in Eq. (23) and with the fabricated sensor's calibration curves; the manuscript's optimization section references such FEM comparison tables, but they are not present, so this check would settle whether the model's symmetry and zero-rotation assumptions actually hold.

Watch

Extended reading notes

Core claim

The paper's central claim is that a non-contact optical six-axis force/torque sensor can be optimized, through a beam-theory model and a chosen matrix objective function, to be both more sensitive and more impact-durable than a commercial capacitive sensor of the same 40 mm form factor. The sensing principle is that three vertical photocouplers measure vertical-load deformation for $F_z$, $M_x$, $M_y$, while three horizontal photocouplers measure horizontal deformation for $F_x$, $F_y$, $M_z$; the T-beam elastomer doubles as a reflective surface, so a single PCB with six ADCs suffices. The authors report a maximum static error of 0.88%, hysteresis below 2%, z-axis resolution more than ten times better than the commercial comparison sensor, and, in the 5000-second gravel test, a maximum drift of 0.375% versus 3.93% for the commercial sensor.

Load-bearing premise

The whole design rests on the assumption that the elastomer deforms as a linear, symmetric Timoshenko beam with equal load sharing among the three arms ($F_z = 3F_C$) and zero rotation at the loading-table connection ($\theta_C = 0$); if impacts or manufacturing tolerances break that symmetry, the optimized dimensions and sensor positions no longer produce the claimed sensitivity and accuracy.

Editorial extensions

If this is right

  • Legged-robot controllers can read foot ground-reaction forces at up to 5 kHz with z-axis resolution roughly an order of magnitude finer than the compared commercial sensor, which matters for balance during fast or uneven-terrain locomotion.
  • The durability result implies that a contactless optical sensing element can survive repeated impacts that permanently shift a capacitive reference sensor, so foot-sensor maintenance intervals could be much longer.
  • Because the optimization method is generic, the same modeling loop can be rerun for other diameter envelopes, force ranges, or elastomer materials without redesigning the sensing principle.
  • A 7th-order polynomial calibration compensates photocoupler nonlinearity with only six ADC channels, making the architecture simpler and cheaper than strain-gauge systems that need external signal processing.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same balance between condition number and sensitivity norms could serve as a general design criterion for any multi-axis sensor whose physics is captured by a linear map from loads to readings.
  • The contactless elastomer may tolerate many more than the tested ten thousand steps; an accelerated cycling-to-failure test would reveal the actual fatigue limit and whether drift stays flat.
  • Because the elastomer doubles as the reflective surface, the design could be scaled down further or embedded directly into joints rather than packaged as a foot-mounted puck.
  • A head-to-head test against a commercial sensor with a matched measurement range would separate the drift advantage due to the contactless principle from the advantage due to the proposed sensor's much larger range.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper presents a non-contact six-axis force/torque sensor for legged robots, using photocouplers to measure deformation of a T-beam elastomer. The authors develop an analytical Timoshenko-beam model, formulate an optimization problem over geometric parameters and sensor positions, and select an objective function based on condition number and norm-based sensitivity criteria. They fabricate an AL7075-T6 prototype, calibrate it with a 7th-order polynomial per channel, and validate it in laboratory static tests and on a custom quadruped robot across posture, speed, and rough-terrain scenarios. The central claim is that the optimized sensor achieves high sensitivity, wide range, and superior durability relative to a commercial RFT40 sensor, with the conclusion citing a maximum drift of 0.375% for the proposed sensor versus 3.93% for the commercial one.

Significance. If the central claims hold, the sensor would be a valuable low-cost, compact alternative for six-axis ground-reaction-force measurement in legged robots, particularly for high-impact applications. The paper does several things well: it explicitly enumerates the optimization objective functions, provides a full six-axis calibration against an ATI MINI-85, reports resolution and hysteresis data, and includes real-robot experiments in three different scenarios. The prototype cost and component count are concrete strengths, and the resolution comparison table (Table VI) gives a transparent basis for the claimed performance advantage. However, the load-bearing durability claim rests on an experiment in which the commercial reference sensor was likely operated beyond its overload rating, and the FEM-based validation of the optimization is missing, which weakens the support for the main conclusions.

major comments (4)
  1. [Section IV-C, Table VIII] The durability comparison is confounded by a mismatch in force ranges. The RFT40 has a z-axis sensing capacity of 150 N with an overload of 450 N, whereas the proposed sensor has a z-axis range of ±1965 N (Table III). The paper itself notes that transient forces on gravel can be several times the ~120 N GRF, so the commercial sensor may have been repeatedly loaded beyond its 450 N overload rating during the ~10,000-step experiment. The observed 3.93% drift therefore does not establish that the proposed sensor is intrinsically more durable; it establishes only that a sensor with roughly 13 times the z-axis range can survive loads that damage a lower-range sensor. Please either match the overload/range conditions of both sensors, record the actual peak forces seen by the RFT40 during the test, or restate the conclusion to acknowledge this confound.
  2. [Section II-B and Section III] There is an unexplained discrepancy between the design target and the reported measurement range. Equation (26) defines the normalization matrix with 520 N and 15.6 N·m as 100%, yet Table III reports a z-axis range of -1965 N to +1965 N, and the conclusion advertises a wide force measurement range. The paper does not explain how the range was extended beyond the 520 N design target, nor whether the optimization constraints in Eq. (25) reflect the larger range. The range claim needs to be reconciled with the optimization formulation.
  3. [Section II-B] The FEM comparison tables are referenced as "table ?? and table ??", but no such tables are present in the manuscript. The comparison of objective functions via FEM is presented as the basis for selecting Cond(G)||G||2/||G||*^2, and Figure 4 is described as the sensor configuration by objective function using FEM. Without the missing FEM tables (including condition numbers and deformation values per objective), the key validation of the analytical model against finite-element results is absent. Please include the missing tables or explicitly state where the data can be found.
  4. [Section II-C, Eqs. (24)-(26)] The sensitivity model is incomplete and internally inconsistent in several places. Equation (24) defines G = R^{-1}_s G using G on both sides, without distinguishing the raw sensitivity matrix from the regulated matrix. Equation (26) specifies R^{-1}_s as a diagonal matrix with entries like "100%/520N", but the position-dependent terms that enter R_s are not defined; for example, how rs2 affects the horizontal sensor channels is not formalized. In Eq. (5), spring constants such as k_{rM xv} and k_{rM xh} appear in the matrix, but Eq. (23) supplies expressions only for a subset of the constants, leaving the remaining entries undefined. These issues make the optimization non-reproducible and should be corrected.
minor comments (4)
  1. [Section II-B] The Timoshenko beam theory is referenced as "[ ?]"; a concrete citation is missing.
  2. [Section II-C] Equation (7) defines theta using the torsional constant It, but the text writes "It = beta*h*b2^3" with beta given as a function that is not formatted clearly; please rewrite the expression and verify the dimensions.
  3. [Section II-B] Equation (27) lists the objective function Cond(G)||G||2/||G||*^2 twice in the last two rows; one of the entries appears to be a duplicate or a typo.
  4. [Section III] Table VI reports RFT40 Fz resolution as 1500 steps with a 0.2 N resolution. If the sensing range is 150 N (as stated in Section IV), the number of steps should be 750; if the range is 300 N, the text should say so explicitly, because the range definition affects the interpretation of the comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the optimization derivation is self-contained and validated against FEM and external sensors.

full rationale

The analytical chain in Sections II.B–II.C derives the compliance matrix G from Timoshenko beam theory with stated assumptions (symmetric three-arm load sharing, zero rotation at C), and each spring constant is computed from beam deflection formulas rather than fitted to the paper's headline results. The 7th-order photocoupler polynomials in Section III are empirical fits, but they convert voltage to displacement and are not reused to select the optimized beam dimensions or detector radii, so no fitted input is relabeled as a prediction. The FEM-based objective-function comparison and the ATI MINI85 / RFT40 validations provide external checks outside the optimization's own G matrix. The self-citations [15], [16] identify the authors' prior photocoupler sensor work; they are disclosed and are not the load-bearing justification for the optimization derivation. The durability comparison in Section IV-C is confounded by the RFT40's lower range (150 N, overload 450 N) and by the proposed sensor being mounted upstream, but this is an experimental validity concern, not a circular derivation. The manuscript does contain two unresolved 'table ??' placeholders where FEM results should appear and a missing reference after 'Timoshenko beam theory' ([ ?]); these are completeness gaps, not circularity. Overall the central optimization is self-contained, so no circular step is identified.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper's central claim rests on the analytical beam model (with several geometric idealizations) and on the calibration polynomials fitted to each photocoupler. The design geometry is optimized rather than fitted to data, but the calibration coefficients are empirical fitting parameters. No new physical entities are introduced.

free parameters (3)
  • Elastomer geometry (l1, l2, b1, b2, h, r, rs2) = l1=14.24, l2=11.00, b1=3.039, b2=0.5535, h=6.690, r=4.437, rs2=8.716 mm for the selected objective Cond(G)||G||2/||G||*2
    These design variables are chosen by the global search optimizer, not measured, and they fully determine the sensor's mechanical response and the claimed optimality.
  • Calibration polynomial coefficients (7th order, one per photocoupler) = not reported in the paper
    Fitted by least squares to map photocoupler voltage to displacement. These coefficients are essential for converting raw readings to forces but are not provided, limiting reproduction.
  • Normalization scale (520 N, 15.6 N·m) = 520 N, 15.6 N·m
    Hand-chosen operating points used in Eq. (26) to normalize the G matrix before optimization. The choice affects the balance of the objective function.
assumptions (4)
  • standard math Timoshenko beam theory is applicable to the T-beam elastomer
    Used to derive the deflection equations in Section II.C; standard solid mechanics, but assumes the beam is slender enough for the theory.
  • domain assumption Symmetric load sharing and zero rotation at the loading table connection
    Eqs. (8)-(9) assume each of the three T-beam arms carries one third of the axial force (Fz = 3FC) and that the connection point C does not rotate. This is a geometric idealization that may fail under asymmetric or impact loading.
  • domain assumption The elastomer remains within its linear elastic range
    All spring constant derivations assume linear stress-strain behavior; the calibration polynomial only corrects photocoupler nonlinearity, not mechanical nonlinearity.
  • domain assumption Photocoupler output is a repeatable function of displacement
    The calibration maps voltage to displacement with a 7th-order polynomial, assuming minimal hysteresis, temperature drift, and aging effects during operation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Parameter Optimization of Optical Six-Axis Force/Torque Sensor for Legged Robots." pith.science (2026). https://pith.science/paper/QETDFCWU

@misc{pith2026250207196,
  author       = {Pith},
  title        = {Pith review of: Parameter Optimization of Optical Six-Axis Force/Torque Sensor for Legged Robots},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QETDFCWU}},
  note         = {Machine review of arXiv:2502.07196}
}
read the original abstract

This paper introduces a novel six-axis force/torque sensor tailored for compact and lightweight legged robots. Unlike traditional strain gauge-based sensors, the proposed non-contact design employs photocouplers, enhancing resistance to physical impacts and reducing damage risk. This approach simplifies manufacturing, lowers costs, and meets the demands of legged robots by combining small size, light weight, and a wide force measurement range. A methodology for optimizing sensor parameters is also presented, focusing on maximizing sensitivity and minimizing error. Precise modeling and analysis of objective functions enabled the derivation of optimal design parameters. The sensor's performance was validated through extensive testing and integration into quadruped robots, demonstrating alignment with theoretical modeling. The sensor's precise measurement capabilities make it suitable for diverse robotic environments, particularly in analyzing interactions between robot feet and the ground. This innovation addresses existing sensor limitations while contributing to advancements in robotics and sensor technology, paving the way for future applications in robotic systems.

Figures

Figures reproduced from arXiv: 2502.07196 by the authors.

Figure 1
Figure 1. a) Configuration and Principle of Proposed Sensor. 3 Vertical Photo [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Elastomer deformation according to the force and moment applied along each axis: (a) Deformation when applying z-axis force ( [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. When deformation occurs along the negative y-axis and the horizontal [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (10 more)
Figure 2
Figure 2. Figure 2: (b) shows the deformation of elastomer when y-axis [PITH_FULL_IMAGE:figures/full_fig_p004_2.png]
Figure 4
Figure 4. Figure 4: Sensor Configuration by Objective Function using FEM [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Sensor Configuration by Objective Function using FEM: (a) Each Part of Sensors and Assembly (b) Calibration Method(Reference Sensor: ATI [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Static Test of Calibration Result (a) Result of x-axis Force (b) Result of y-axis Force (c) Result of z-axis Force (d) Result of x-axis Moment (e) [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: x-axis Force Repeatability Test Result during 100s. [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Simultaneous Force and Moment Applying Experiment(Crosstalk Test Result) (a) Result of x-axis Force (b) Result of y-axis Force (c) Result of z-axis [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: Application of Proposed Sensor and Experiments for Verification of Proposed Sensor (a) A Picture of the foot with sensors (b) Posture Adjustment [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Postural Adjustment Experiment of A Quadruped Robot’s Torso (a) Position Graph of Torso (b) Rotation Graph of Torso (c) Ground Reaction Force [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 11
Figure 11. Figure 11: Speed Adjustment Experiment of A Quadruped Robot Result (a) Left Foot Ground Reaction Force at 0.2m/s (b) Left Foot Ground Reaction Force [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: Rough Terrain Experiment Result(Durability Test) (a) Result of Left [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. A Miniature High-Resolution Tension Sensor Based on a Photo-Reflector for Robotic Hands and Grippers

    cs.RO 2025-07 conditional novelty 5.0 of 10

    A photo-reflector-based miniature tension sensor with a symmetric elastomer and flexure hinges achieves ~9.9 mN resolution and 0.455 N RMSE, but the 200 N full-scale claims rest on calibration data limited to 70 N.

Reference graph

Works this paper leans on

21 extracted references · 19 canonical work pages · cited by 1 Pith paper

  1. [15]

    A compact six-axis force/torque sensor using photo- couplers for impact robustness,

    H.-B. Kim et al., “A compact six-axis force/torque sensor using photo- couplers for impact robustness,”Review of Scientific Instruments, vol. 95, no. 4, 2024

    work page 2024

  2. [16]

    A compact optical six-axis force/torque sensor for legged robots using a polymorphic calibration method,

    H.-B. Kim et al. , “A compact optical six-axis force/torque sensor for legged robots using a polymorphic calibration method,” arXiv preprint arXiv:2309.04720, 2023

    arXiv 2023

  3. [1]

    Dynamic locomotion in the mit cheetah 3 through convex model-predictive con- trol,

    J. Di Carlo, P. M. Wensing, B. Katz, G. Bledt, and S. Kim, “Dynamic locomotion in the mit cheetah 3 through convex model-predictive con- trol,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , IEEE, 2018, pp. 1–9

    work page 2018

  4. [2]

    Mini cheetah: A platform for pushing the limits of dynamic quadruped control,

    B. Katz, J. Di Carlo, and S. Kim, “Mini cheetah: A platform for pushing the limits of dynamic quadruped control,” in 2019 International Conference on Robotics and Automation (ICRA), IEEE, 2019, pp. 6295– 6301

    work page 2019

  5. [3]

    Anymal-a highly mobile and dynamic quadrupedal robot,

    M. Hutter et al. , “Anymal-a highly mobile and dynamic quadrupedal robot,” in 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , IEEE, 2016, pp. 38–44

    work page 2016

  6. [4]

    Design of kaist hound, a quadruped robot platform for fast and efficient locomotion with mixed-integer nonlinear optimization of a gear train,

    Y .-H. Shin et al., “Design of kaist hound, a quadruped robot platform for fast and efficient locomotion with mixed-integer nonlinear optimization of a gear train,” in 2022 International Conference on Robotics and Automation (ICRA), IEEE, 2022, pp. 6614–6620

    work page 2022

  7. [5]

    Learning quadrupedal locomotion on deformable terrain,

    S. Choi et al., “Learning quadrupedal locomotion on deformable terrain,” Science Robotics, vol. 8, no. 74, p. eade2256, 2023

    work page 2023

  8. [6]

    Multi-axis force-torque sensors for measuring zero-moment point in humanoid robots: A review,

    J.-H. Kim, “Multi-axis force-torque sensors for measuring zero-moment point in humanoid robots: A review,” IEEE Sensors Journal , vol. 20, no. 3, pp. 1126–1141, 2019

    work page 2019

Show all 21 references
  1. [7]

    A surgical palpation probe with 6-axis force/torque sensing capability for minimally invasive surgery,

    U. Kim et al. , “A surgical palpation probe with 6-axis force/torque sensing capability for minimally invasive surgery,” IEEE Transactions on Industrial Electronics , vol. 65, no. 3, pp. 2755–2765, 2017

    2017

  2. [8]

    The msc prosthetic hand: Rapid, powerful, and intuitive,

    Y . Cho et al., “The msc prosthetic hand: Rapid, powerful, and intuitive,” IEEE Robotics and Automation Letters , vol. 7, no. 2, pp. 3170–3177, 2022

    2022

  3. [9]

    Design of a miniature force sensor based on photointerrupter for robotic hand,

    S. H. Jeong et al. , “Design of a miniature force sensor based on photointerrupter for robotic hand,” Sensors and Actuators A: Physical , vol. 269, pp. 444–453, 2018

    2018

  4. [10]

    Design of a strain gauge-based force sensor with three ranges based on compliant mechanisms,

    F. Chen et al. , “Design of a strain gauge-based force sensor with three ranges based on compliant mechanisms,” IEEE Transactions on Instrumentation and Measurement , 2025

    2025

  5. [11]

    A novel six-axis force/torque sensor for robotic appli- cations,

    U. Kim et al. , “A novel six-axis force/torque sensor for robotic appli- cations,” IEEE/ASME Transactions on Mechatronics, vol. 22, no. 3, pp. 1381–1391, 2016

    2016

  6. [12]

    Bi-modal hemispherical sensors for dynamic locomo- tion and manipulation,

    L. Epstein et al., “Bi-modal hemispherical sensors for dynamic locomo- tion and manipulation,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , IEEE, 2020, pp. 7381–7381

    2020

  7. [13]

    A compact two dof magneto-elastomeric force sensor for a running quadruped,

    A. Ananthanarayanan et al. , “A compact two dof magneto-elastomeric force sensor for a running quadruped,” in 2012 IEEE International Conference on Robotics and Automation , IEEE, 2012, pp. 1398–1403

    2012

  8. [14]

    Six-dimensional force/torque sensor based on fiber bragg gratings with low coupling,

    L. Xiong et al. , “Six-dimensional force/torque sensor based on fiber bragg gratings with low coupling,” IEEE Transactions on Industrial Electronics, vol. 68, no. 5, pp. 4079–4089, 2020

    2020

  9. [17]

    Robotous - innovative robotic solutions,

    “Robotous - innovative robotic solutions,” http://www.robotous.com/ main, accessed: 2025-01-21

    2025

  10. [18]

    Ati industrial automation,

    “Ati industrial automation,” https://www.ati-ia.com/, accessed: 2025-01- 21

    2025

  11. [19]

    Highly dynamic quadruped locomotion via whole- body impulse control and model predictive control,

    D. Kim et al. , “Highly dynamic quadruped locomotion via whole- body impulse control and model predictive control,” arXiv preprint arXiv:1909.06586, 2019

    1909 arXiv

  12. [20]

    LCM: Lightweight communications and mar- shalling,

    A. S. Huang et al. , “LCM: Lightweight communications and mar- shalling,” in 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems , IEEE, 2010, pp. 4057–4062. ARXIV , JANUARY 2025 14 VI. B IOGRAPHY SECTION Hyun-Bin Kim received the B.S., M.S. and Ph.D. degr...

    2010

  13. [2005]

    Since 2007, he has been with the Department of Mechanical Engineering, KAIST

    In 2005, he joined the Department of Me- chanical Engineering, Korea Polytechnic University, Siheung, Republic of Korea, as a Faculty Member. Since 2007, he has been with the Department of Mechanical Engineering, KAIST. His research interests include control theory, electric v...

    2005

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.