Loop Shaping of Hybrid Motion Control with Contact Transition

Michael Ruderman

REVIEW 3 major objections 7 minor 24 references

Loop Shaping of Hybrid Motion Control with Contact Transition

T0 review · 3 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Loop shaping of the disturbance sensitivity function $S(s)=1/(Z(s)s)$ lets a stiff displacement-feedback controller reconfigure on contact without any force measurement, as demonstrated on a 1-DOF actuator meeting a grape.

desk verdict A clean proof-of-concept for sensor-free contact-adaptive loop shaping, undermined by a hand-tuned threshold and missing switched-system analysis; worth reviewing, not worth taking as a guarantee. read the letter →

arxiv 2411.19495 v2 pith:VQNCZCOH submitted 2024-11-29 eess.SY cs.ROcs.SY

Michael Ruderman This is my paper · ORCID

classification eess.SYcs.ROcs.SY MSC 93B5293C80

keywords motioncontrolcontacttransitionimpedanceloopshapingdisturbancesensitivityfunctionsensor-freeswitchingvoice-coilactuatorsoftenvironment

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 shows that an ordinary stiff displacement-feedback controller, the kind that usually makes a motion axis push straight through anything it meets, can be turned into a hybrid controller that gives way smoothly on contact, and that this can be done without a force sensor. The design works by shaping the disturbance sensitivity function $S(s)$ — the transfer function from contact force to displacement deviation — to match the environment's impedance $Z(s)$ through $S(s)=1/(Z(s)s)$. The switch from stiff to compliant control is triggered simply by the control signal exceeding a preset bound $U$, so the only measured quantity remains the output displacement. Experiments on a 1-DOF voice-coil actuator with a grape as the soft, penetrable object show the stiff controller ploughing into the fruit while the reshaped controllers either repel it or press it with bounded force.

What carries the argument

The load-bearing identity is $S(s)=1/(Z(s)s)$, which ties the closed-loop disturbance sensitivity to the contact environment's impedance and follows from comparing the control stiffness definition, the impedance ratio, and the sensitivity transfer function. The design formulas that carry the argument are the stiff controller $C_s(s)=\omega_0^2/(G(s)s(s+2\omega_0))$ obtained from a critically damped reference model, and the reshaped viscous controller $C_v(s)=(\alpha s G(s)-1)/G(s)$, with a low-pass filter added for properness; the viscoelastic controller $C_{ve}$ adds a saturated proportional error term to $C_v$. The switching law is a threshold comparison $|u(t)|>U$ on the already-available control signal, supported by the disturbance-to-control transfer function which shows that an over-limit control value indicates an external force. Together these pieces make the contact transition detectable and executable with no force measurement.

What would settle it

Run the same no-contact reference trajectory with added Coulomb friction or increased measurement noise so that $|u(t)|$ exceeds $U=1.3$ at some instant; if the controller switches to the soft mode with no object present, the sensor-free contact detection is falsified. Equivalently, a contact that is so soft that $|u|$ never exceeds $U$ would not trigger the switch, and the tool would penetrate.

Watch

Extended reading notes

Core claim

The paper's central claim is that the disturbance sensitivity function $S(s)=x(s)/F(s)$ of a closed-loop motion system is the inverse of the control stiffness operator, so prescribing an environmental impedance $Z(s)$ is equivalent to prescribing $S(s)=1/(Z(s)s)$. From this identity, the author derives a stiff controller $C_s$ given by a critically damped reference model, and two reshaped controllers, a purely viscous $C_v$ and a viscoelastic $C_{ve}$, that realize the desired contact impedance. The hybrid scheme switches from $C_s$ to $C_v$ or $C_{ve}$ whenever the magnitude of the control signal exceeds a threshold $U$, justified through the disturbance-to-control transfer function $U(s)=u(s)/F(s)$. In the experimental case study, the stiff controller drives the tool into the grape, whereas the viscous controller produces a slightly repulsive response and the viscoelastic controller presses the grape with a bounded, saturated force while avoiding penetration.

Load-bearing premise

The load-bearing premise is that in contact-free motion the control signal stays below the chosen threshold $U$, so $|u(t)|>U$ uniquely indicates contact; the paper sets $U=1.3$ and shows one no-contact trajectory, but does not analyze how friction, measurement noise, or trajectory transients affect this margin.

Editorial extensions

If this is right

Any position-controlled axis with a known nominal plant $G(s)$ of relative degree at most 2 can be extended with a safe contact mode by reshaping $S(s)$, without adding hardware.
Contact detection reduces to a threshold test on the control signal, so the method runs on existing real-time controllers that already log $u(t)$.
The two reshaped controllers give distinct behaviors — viscous repulsion or viscoelastic bounded-force pressing — so the same switching law can be tuned to different tasks by choosing $C_v$ or $C_{ve}$.
Because the design is carried out in the frequency domain, the non-contact performance can still be analyzed with standard Bode and sensitivity tools.
The experimental grape scenario suggests applicability to medical palpation and fragile-object handling, where penetration must be avoided without a dedicated force sensor.

Reading between the lines

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

The threshold $U$ is the only ad hoc parameter in the method; a systematic margin analysis that includes Coulomb friction, measurement noise, and trajectory transients would make the switching decision provably robust.
The identity $S(s)=1/(Z(s)s)$ invites generalization to nonlinear or Hunt-Crossley contact models by shaping $S(s)$ accordingly, although the paper only treats linear viscous and viscoelastic environments.
Near the threshold, the switching law could chatter if the contact force hovers around $U$; adding hysteresis or a dwell time is an obvious practical hardening.
After switching, the post-contact control signal encodes information about the environment's impedance, suggesting a force-free way to estimate tissue or material softness from the same measurements already used for control.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

You should know this is a short, honest proof-of-concept paper, not a new theory. The genuinely new combination is loop-shaping the disturbance sensitivity to match an assumed environment impedance and then using the control signal itself, rather than a force sensor, as the contact detector. The algebra around S(s)=1/(Z(s)s) is elementary, since it falls out of the definitions, but it gives a clean design target, and the authors are not pretending otherwise. The Bode reasoning is correct, the reshaped controllers are made proper with low-pass filters, and the grape experiment is a real demonstration: the stiff controller ploughs into the grape while the hybrid controllers hold position or gently repel. The two linked videos are a nice extra. Self-citations are mostly to prior hardware and friction work; they do not carry the core claim, so I do not see a citation-pattern problem. The soft spots are real but not fatal. The load-bearing assumption is the threshold U=1.3, set by hand from a single no-contact run. Figure 7 shows friction transients and measurement noise, and the paper gives no bound on the worst-case |u(t)| during free motion. If a transient pushes the control signal above U, the system switches without contact; if the contact is very soft, the switch may come only after noticeable penetration. That directly conflicts with the intro's phrase that a smooth and stable contact transition can be guaranteed. The paper also gives no stability or boundedness analysis for the switched system and does not discuss controller states at the moment of switching. The mapping from the design parameters alpha and omega_c to the numerical coefficients in (15)-(16) is not shown, which weakens reproducibility. The experiments are single-run and qualitative: no repeated trials, no contact-force ground truth, just displacement and control traces. Each of these is more a missing-evidence problem than a contradiction in the derivation. The reader's conditional verdict matches mine. The stress-test concern about the threshold margin does land on reading the paper, but it should be a review request, not a rejection. This paper is for control engineers working on soft-contact motion, especially in medical or service robotics, and it would be a good reading-group example of impedance versus admittance causality done with standard frequency-domain tools. I would send it to peer review rather than desk reject; a good referee would ask for a threshold-margin analysis and some argument about the switched system. I would cite the revised version if the threshold procedure and switched dynamics get addressed; as it stands, my own verdict is conditional.

Referee Report

3 major / 7 minor

Summary. The paper proposes a hybrid motion control scheme that switches from a stiff displacement-feedback controller to a compliant (viscous or viscoelastic) controller when the magnitude of the control signal exceeds a fixed threshold U, thereby detecting contact without a force sensor. The design is based on loop shaping of the disturbance sensitivity function S(s), with the target S_v(s)=1/(alpha s) for viscous contact and a low-pass-filtered version for implementation. The approach is demonstrated on a 1-DOF voice-coil actuator whose tool contacts a half grape, comparing stiff PID, hybrid Cs->Cv, and hybrid Cs->Cve. The paper claims smooth and stable contact transition using only the measured displacement and control signal.

Significance. The contribution is a concise, frequency-domain interpretation of impedance control that requires no force measurement and no state observer. The experimental demonstration with soft, penetrable objects is valuable, and the controller equations are internally consistent for the nominal LTI plant. However, the paper does not provide a stability analysis of the switched system, does not quantify the robustness of the contact-detection threshold, and does not show how the design parameters alpha and omega_c map to the experimental controller coefficients, so the generality of the claimed guarantee exceeds what is demonstrated. The paper is a useful proof-of-concept rather than a fully supported design method.

major comments (3)

[Section IV-B, Fig. 7] The contact detection relies entirely on the fixed threshold U=1.3. The paper gives no worst-case bound for |u(t)| during contact-free motion; Fig. 7 itself shows friction-induced transients after slope segments, so it is plausible that some other trajectory or a slightly larger disturbance would push |u| above U without contact, causing false switching, or that a soft-contact force would not push |u| above U before unacceptable penetration. Since no force sensor is present to correct the decision, this gap is load-bearing for the sensor-free switching claim.
[Section IV-B, Eqs. (15)-(16)] The controllers in (15)-(16) are presented with numerical coefficients, but the paper does not show how these coefficients follow from the design parameters alpha and omega_c plus the plant parameters K and tau in (13). For instance, with G(s)=K/(s(tau s+1)) and K=0.0408, tau=0.00668, the expression C_v(s)=alpha s - 1/G(s) has particular polynomial coefficients; the coefficients in (15) differ from a direct substitution, and the paper does not display the intermediate derivation or the chosen values of alpha and omega_c. Providing this mapping is necessary for the design to be reproducible.
[Section III and Section IV-B] The paper claims that the contact transition is 'smooth and stable' and that the hybrid control 'guarantees' stable transitions, but no stability analysis of the switched system is given. The switching condition depends on the state-dependent control signal |u(t)|, and the initialization of the controller states (the integrator in (14) and the filter states in (15)-(16)) at the switch is not specified. The absence of this analysis is a central weakness for a paper whose main claim is a stable contact transition; at least a per-mode stability check and a discussion of chattering avoidance are required.

minor comments (7)

[Section I, title] The title on the first page shows 'Transiti on' with a spacing error; it should read 'Transition'.
[Section II] The word 'senors' should be 'sensors'.
[Section IV-B] The phrase 'The first one, denoted by Cv(s), is enabling a purely viscous...' should use 'enables' instead of 'is enabling'.
[Fig. 5 caption] The phrase 'shown over each other' is informal; use 'superimposed'.
[Section III] The phrase 'an overshot |u(t)| > U' should read 'an overshoot |u(t)| > U'.
[Section IV-B, Eq. (16)] The saturation function sat_U[kp e(t)] is not formally defined; the argument and output should be specified precisely.
[Section III and IV-B] The paper should state whether the switching logic includes hysteresis or a minimum dwell time, since the same threshold U is used for both activation and deactivation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the loop-shaping derivation is self-contained and the experimental controllers are solved from stated sensitivity targets, not fitted to the contact outcome.

full rationale

The central derivation is algebraic and non-circular. Starting from S(s)=G/(1+CG) and the stated identity S=1/(Zs), the controller solving C=(ZsG-1)/G is obtained by block-diagram algebra; the paper is explicit that this is the design equation, and the target sensitivity S_v=1/(alpha*s) is chosen a priori from the viscous-environment model Z_v=alpha, not from the contact data. The plant parameters K and tau are identified in closed loop and confirmed by the Bode plot in Fig. 5, so the subsequent controllers are not reverse-engineered from the contact response. The switching threshold U=1.3 is selected as a nominal bound from the no-contact run and used as a design parameter; it is not a fitted output presented as a prediction, and the grape-contact runs are independent demonstrations. References to the author's previous work ([13], [19], [22], [23], [24]) provide identification details, friction models, and an existing robust PID design, but the load-bearing loop-shaping relation, the controller expressions, and the experimental evidence are contained in this paper, so the self-citations do not carry the central claim. No self-definitional step, fitted-input-as-prediction, uniqueness import, or ansatz-by-citation pattern is present. The absence of a formal switched-system stability proof is a correctness/robustness concern, not a circularity concern.

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

The method relies on a hand-set contact threshold, an assumed environment impedance class, and a nominal LTI plant model. No new physical entities are introduced.

free parameters (3)

Contact detection threshold U = 1.3 V
Set by hand in Section IV.B; the sensor-free switching depends on this bound, but no tuning rule or robustness margin is provided.
Viscous coefficient alpha in S_v(s)=1/(alpha s) = 100 in the design example; experimental value implicit in the coefficients of (15) and not stated
Defines the desired contact admittance; different alpha changes the repulsive behavior. In the experiments the corresponding coefficients in (15)-(16) are given numerically but alpha itself is not reported.
Low-pass filter cutoff omega_c in (15)-(16) = Not specified, described only as 'sufficiently high'
Needed to make the improper controllers (12) realizable; its value affects the achievable bandwidth and noise amplification, and it is not reported.

assumptions (4)

domain assumption The plant is accurately described by G(s)=K/[s(tau s+1)] with K=0.0408 and tau=0.00668, neglecting electrical time constant and delay.
Used in Section IV.A for the design; neglected dynamics are only mentioned as a robustness challenge, not analyzed.
domain assumption The contact environment is a linear time-invariant impedance belonging to the classes in Section II, namely viscous alpha or Kelvin-Voigt alpha+beta/s.
The design of S_v and S_ve assumes this; a real grape is nonlinear, and no identification of its impedance is performed.
domain assumption The disturbance force F(t) enters as a matched input at the plant input, so that u+F is the total force acting on the plant.
Standard in the motion control loop in Section II, but contact forces act at the tool tip, not exactly at the motor input; the mapping between the two is not discussed.
ad hoc to paper The control signal threshold U can distinguish contact from friction, measurement noise, and tracking transients.
The sensor-free detection in Section III relies on this; the paper does not prove or measure a margin for this bound.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Loop Shaping of Hybrid Motion Control with Contact Transition." pith.science (2026). https://pith.science/paper/VQNCZCOH

@misc{pith2026241119495,
  author       = {Pith},
  title        = {Pith review of: Loop Shaping of Hybrid Motion Control with Contact Transition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VQNCZCOH}},
  note         = {Machine review of arXiv:2411.19495}
}

read the original abstract

A standard motion control with feedback of the output displacement cannot handle unforeseen contact with environment without penetrating into the soft, i.e. viscoelastic, materials or even damaging the fragile materials. Robotics and mechatronics with tactile and haptic capabilities, and in particular medical robotics for example, place special demands on the advanced motion control systems that should enable the safe and harmless contact transitions. This paper shows how the basic principles of loop shaping can be easily used to handle sufficiently stiff motion control in such a way that it is extended by sensor-free dynamic reconfiguration upon contact with the environment. A thereupon based hybrid control scheme is proposed. A remarkable feature of the developed approach is that no measurement of the contact force is required and the input signal and the measured output displacement are the only quantities used for design and operation. Experiments on 1-DOF actuator are shown, where the moving tool comes into contact with grapes that are soft and simultaneously penetrable.

Figures

Figures reproduced from arXiv: 2411.19495 by the authors.

**Figure 2.** Magnitude response of G(jω), H(jω), Ss(jω), and Sv(jω). from the Bode plot of Ss(jω) that a step-wise disturbance F(s), which appears at the contact instant, will be largely suppressed by the stiff motion control. Following to that, the controlled mechanical motion system will continue to move and penetrate into the environmental object, or moves it away from itself if the latter is not fixed. Quite the opposite, th… view at source ↗

**Figure 3.** Magnitude response of U(jω) for Cs and Cv controllers. can recognize that since r is set to zero for Cv, and the control magnitude response of both Cs and Cv have nearly the same value, an appearance of the step-wise disturbance F(tc) at t = tc will lead to a decrease of |u(tc)| = U by the magnitude equal to |F| for t > tc. Then, the force imposed on the environmental object under contact drops respectively. IV. EXP… view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Experimental setup of the controlled motion in conta [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

**Figure 7.** Figure 7: Control response without contacting environment: ( [PITH_FULL_IMAGE:figures/full_fig_p005_7.png]

**Figure 6.** Figure 6: Magnitude response of the disturbance sensitivity fu [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]

**Figure 8.** Figure 8: Control response with contacting environment, comp [PITH_FULL_IMAGE:figures/full_fig_p006_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 24 canonical work pages

[1]

Dynamic stability issues in for ce control of manipulators,

C. An and J. Hollerbach, “Dynamic stability issues in for ce control of manipulators,” in American Control Conference , 1987, pp. 821–827

work page 1987
[2]

Impedance control: An approach to manipulati on: Part I – theory,

N. Hogan, “Impedance control: An approach to manipulati on: Part I – theory,” Transactions of the ASME , vol. 107, no. 1, pp. 1–7, 1985

work page 1985
[3]

Hybrid impedance control of ro botic manipulators,

R. Anderson and M. Spong, “Hybrid impedance control of ro botic manipulators,” IEEE Journal on Robotics and Automation , vol. 4, no. 5, pp. 549–556, 1988

work page 1988
[4]

Controlling contact transitio n,

J. Hyde and M. Cutkosky, “Controlling contact transitio n,” IEEE Control Systems Magazine , vol. 14, no. 1, pp. 25–30, 1994

work page 1994
[5]

Force control of robot manipulators,

T. Y oshikawa, “Force control of robot manipulators,” in IEEE Inter- national Conference on Robotics and Automation , 2000, pp. 220–226

work page 2000
[6]

Force control,

L. Villani and J. De Schutter, “Force control,” in Springer Handbook of Robotics , pp. 195–220, 2016

work page 2016
[7]

Estimation, identiﬁ cation, and sensorless control in motion control system,

K. Ohnishi, N. Matsui, and Y . Hori, “Estimation, identiﬁ cation, and sensorless control in motion control system,” Proceedings of the IEEE , vol. 82, no. 8, pp. 1253–1265, 1994

work page 1994
[8]

A uniﬁed pa ssivity- based control framework for position, torque and impedance control of ﬂexible joint robots,

A. Albu-Sch¨ affer, C. Ott, and G. Hirzinger, “A uniﬁed pa ssivity- based control framework for position, torque and impedance control of ﬂexible joint robots,” The International Journal of Robotics Research , vol. 26, no. 1, pp. 23–39, 2007

work page 2007

Show all 24 references

[9]

Hybrid position/force control of manipula- tors,

M. Raibert and J. Craig, “Hybrid position/force control of manipula- tors,” ASME Journal of Dynamic Systems, Measurement, and Control , vol. 103, no. 2, pp. 126–133, 1981

1981
[10]

V ariable impedance control of redundant manipulators for intuitive human–robot physi cal interac- tion,

F. Ficuciello, L. Villani, and B. Siciliano, “V ariable impedance control of redundant manipulators for intuitive human–robot physi cal interac- tion,” IEEE Tran. on Robotics , vol. 31, no. 4, pp. 850–863, 2015

2015
[11]

Hybrid position/force con trol for hy- draulic actuators,

P . Pasolli and M. Ruderman, “Hybrid position/force con trol for hy- draulic actuators,” in IEEE 28th Mediterranean Conference on Control and Automation (MED) , 2020, pp. 73–78

2020
[12]

Basic problems in stabilit y and design of switched systems,

D. Liberzon and A. S. Morse, “Basic problems in stabilit y and design of switched systems,” IEEE Control Systems Magazine , vol. 19, pp. 59–70, 1999

1999
[13]

On switching between motion and force con trol,

M. Ruderman, “On switching between motion and force con trol,” in IEEE 27th Mediterranean Conference on Control and Automati on (MED), 2019, pp. 445–450

2019
[14]

Accuracy/robustness dilemma in impedance control,

T. V alency and M. Zacksenhouse, “Accuracy/robustness dilemma in impedance control,” ASME Journal of dynamic systems, measurement, and control, vol. 125, no. 3, pp. 310–319, 2003

2003
[15]

A hybrid system f ramework for uniﬁed impedance and admittance control,

C. Ott, R. Mukherjee, and Y . Nakamura, “A hybrid system f ramework for uniﬁed impedance and admittance control,” Journal of Intelligent & Robotic Systems , vol. 78, pp. 359–375, 2015

2015
[16]

Motion-contr ol tech- niques of today and tomorrow: a review and discussion of the c hal- lenges of controlled motion,

M. Ruderman, M. Iwasaki, and W.-H. Chen, “Motion-contr ol tech- niques of today and tomorrow: a review and discussion of the c hal- lenges of controlled motion,” IEEE Industrial Electronics Magazine , vol. 14, no. 1, pp. 41–55, 2020

2020
[17]

Spong, S

M. Spong, S. Hutchinson, and M. Vidyasagar, Robot Modeling and Control. John Wiley & Sons, 2005

2005
[18]

K. J. ˚Astr¨ om and R. Murray, Feedback systems: an introduction for scientists and engineers . Princeton University Press, 2021

2021
[19]

Ruderman, Analysis and compensation of kinetic friction in robotic and mechatronic control systems

M. Ruderman, Analysis and compensation of kinetic friction in robotic and mechatronic control systems . CRC Press, 2023

2023
[20]

G uaranteeing prescribed performance and contact maintenance via an appr oximation free robot force/position controller,

C. P . Bechlioulis, Z. Doulgeri, and G. A. Rovithakis, “G uaranteeing prescribed performance and contact maintenance via an appr oximation free robot force/position controller,” Automatica, vol. 48, no. 2, pp. 360–365, 2012

2012
[21]

Collision detection and safe reaction with the DLR-III lig htweight manipulator arm,

A. De Luca, A. Albu-Schaffer, S. Haddadin, and G. Hirzin ger, “Collision detection and safe reaction with the DLR-III lig htweight manipulator arm,” in 2006 IEEE/RSJ International Conference on Intelligent Robots and Systems , 2006, pp. 1623–1630

2006
[22]

Motion control with optimal nonlinear da mping: from theory to experiment,

M. Ruderman, “Motion control with optimal nonlinear da mping: from theory to experiment,” Control Engineering Practice , vol. 127, p. 105310, 2022

2022
[23]

Dis turbance sensitivity analysis and experimental evaluation of conti nuous sliding mode control,

M. Ruderman, J. Reger, B. Calmbach, and L. Fridman, “Dis turbance sensitivity analysis and experimental evaluation of conti nuous sliding mode control,” arXiv preprint arXiv:2208.06608 , 2022

2022 arXiv
[24]

On convergence analysis of feedback cont rol with inte- gral action and discontinuous relay perturbation,

M. Ruderman, “On convergence analysis of feedback cont rol with inte- gral action and discontinuous relay perturbation,” Comm. in Nonlinear Science and Numerical Simulation , vol. 145, p. 108698, 2025

2025

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Dynamic stability issues in for ce control of manipulators,

C. An and J. Hollerbach, “Dynamic stability issues in for ce control of manipulators,” in American Control Conference , 1987, pp. 821–827

work page 1987

[2] [2]

Impedance control: An approach to manipulati on: Part I – theory,

N. Hogan, “Impedance control: An approach to manipulati on: Part I – theory,” Transactions of the ASME , vol. 107, no. 1, pp. 1–7, 1985

work page 1985

[3] [3]

Hybrid impedance control of ro botic manipulators,

R. Anderson and M. Spong, “Hybrid impedance control of ro botic manipulators,” IEEE Journal on Robotics and Automation , vol. 4, no. 5, pp. 549–556, 1988

work page 1988

[4] [4]

Controlling contact transitio n,

J. Hyde and M. Cutkosky, “Controlling contact transitio n,” IEEE Control Systems Magazine , vol. 14, no. 1, pp. 25–30, 1994

work page 1994

[5] [5]

Force control of robot manipulators,

T. Y oshikawa, “Force control of robot manipulators,” in IEEE Inter- national Conference on Robotics and Automation , 2000, pp. 220–226

work page 2000

[6] [6]

Force control,

L. Villani and J. De Schutter, “Force control,” in Springer Handbook of Robotics , pp. 195–220, 2016

work page 2016

[7] [7]

Estimation, identiﬁ cation, and sensorless control in motion control system,

K. Ohnishi, N. Matsui, and Y . Hori, “Estimation, identiﬁ cation, and sensorless control in motion control system,” Proceedings of the IEEE , vol. 82, no. 8, pp. 1253–1265, 1994

work page 1994

[8] [8]

A uniﬁed pa ssivity- based control framework for position, torque and impedance control of ﬂexible joint robots,

A. Albu-Sch¨ affer, C. Ott, and G. Hirzinger, “A uniﬁed pa ssivity- based control framework for position, torque and impedance control of ﬂexible joint robots,” The International Journal of Robotics Research , vol. 26, no. 1, pp. 23–39, 2007

work page 2007