Thesis work

Design of a Stable and Transparent Micro Teleoperation System

Advisors : Pr. Stéphane Régnier and Pr. Vincent Hayward.

Defended on 5 th March 2013 at ISIR.

Electronic version of the thesis is available on tel.archives (link)

Video : Micro teleoperation System (user interaction with a water droplet) [Youtube video]

Keywords : microrobotics, teleoperation, haptic interface, force measurement, electrostatic actuator, coupling scheme,transparency, stability.

Direct human interaction with the microscale world is an important challenge in microrobotics. Such direct interaction has applications in nano-technology, biology,material characterization, or in the investigation of the physics of microscale phenomena. In these situations, operators are confronted with manual tasks that are highly unusual and unpredictable under impoverished sensory conditions. Direct interaction with the microscale is nowadays typically performed while viewing the task through a microscope -- frequently with no other depth cue than blur and focus -- and with the motion of objects that are governed by unusual physics. Mechanical behavior is no longer dominated by gravity and elasticity, and short range forces including electrostatic, capillary, and van der Waals forces dominate. Force-feedback teleoperation with a sufficient level of fidelity in the reproduction of these physics would assist the operators incarrying out manipulation and assembly tasks, among other options.

Since the 1990s, several such micro teleoperation systems have been proposed with a view to provide haptic feedback in order to improve manipulation capabilities at the microscale. However, an essential shortcoming of these systems is a lack of transparency. The force signal fed back to the user is significantly different from that taking place in microscopic physical phenomenon. The causes of this problem are threefold. The first is computational, a micro-to-macro teleoperation chain requires scaling gains in the order of 10^4 to 10^7. With those values, it is difficult to simultaneously guarantee stability and transparency of the micro teleoperation system. The second cause of loss of transparency is the probe itself, which is used to interact with the sample. The most common type is based on the deflection of an elastic cantilever as in Atomic force Microscopy (AFM), where the interaction force is measured by monitoring the deflection of the cantilever. Elastic elements, however, introduce instabilities such as a jump to contact when the cantilever tip is brought close to the sample where the gradient of the attraction force exceeds the cantilever stiffness. As a result, the deflection of the probe is no longer is an accurate reflection of the interaction force. The third cause comes from the force feedback devices. They are ill-fitted to render the high dynamic range typical of microscale phenomena as well as commercially available devices.

To tacle those limitations, this thesis proposed to construct a micro teleoperation chain for microscale applications that preserves stability as well as transparency. Both the master and slave devices were especially designed to achieve this purpose and to enable the application of a direct control scheme. In direct coupling, the position of the haptic device is scaled down so as to serve as set-point for the probe position relative to the sample, and the measured force is scaled up to be fed back to the user. The stability of such a system can be established when both the master and slave are passive. It follows that each component of the system should individually be designed and controlled to exhibit passive linear behavior. In sum, the several issues were addressed in this purpose :

  • design of an adapted haptic device,
  • design of a force sensor with proper dynamic properties and bandwidth,
  • development of a case-specific control scheme,
  • implementation of bilateral coupling between the haptic device and the force sensor.

Haptic device

Haptic interfaces based on force feedback enable bidirectional human-system interaction through the sens of touch in response to user movement. However, haptic interfaces using electromagnetic drives are subjected to an inherent inertia-torque trade-off which limits transparency : the higher the torque, the higher the inertia. Consequently the transparency is degraded with inertia. In order to cope with this limitation, we used in this work a dual-stage haptic interface introduced by [Millet et al. 09].

The haptic interface used in this work is based on a dual-stage architecture. It is comprised of a large motor coupled to a small one, also carrying the handle, through a passive viscous clutch based on eddy currents. This clutch mechanism accurately transforms slip velocity of the large motor into torque and exhibits a linear behavior on a large bandwidth. This first order relationship between the velocity control and the handle eliminates the dynamics of the large motor from the user experience. A feed-forward path is provided through the small motor to fill in the transients or fast variations of the input signal that the large motor could not provide. As this second motor has negligible friction and inertia, unwanted forces could be kept below human detection threshold achieving quasi-perfect transparency.

Micro-force sensor

The slave device performs the physical interaction at the microscale and provides force sensing. The force sensor designed in this work is presented in depth in [abdenbi et al. 13]. It uses electrostatic energy instead of elastic energy to perform the measurement. It is designed around a linear comb-drive actuator (see Fig.2). Its working principle is to keep the probe on its equilibrium position by a force to cancel out external forces. As the comb-drive actuator is perfectly known and controlled, the balancing force is equal to the external force and directly provides the measure. Based on the position feedback of the probe, an H-infinity controller generates an opposite signal in order to balance the external applied force and to maintain the probe at its equilibrium position.

Instead of measuring the interaction using the signal coming from the displacement of the probe, it is computed as being proportional to the value of the opposite voltage applied by the controller generating the opposite force. In this case, force developed by the sensor is equal to the external force. This technique which uses the closed-loop control principle allows maintaining the sensitive part of the sensor in the linear range, improving precision, increasing the sensor dynamic range and improving the robustness and the stability of the sensor. Thus, the sensor probe can track the integrality of the tip-sample interaction curves. It especially avoids several problems inherent to scanning probe microscopy using elastic probes, such as pull-in, pull-off and hysteresis.


Fig.2 : Active force sensor (slave device) Fig.3 : Micro teleoperation system.

Micro teleoperation chain

A direct force feedback control scheme, Fig.3, requiring two homothetic gains is used to connect the haptic device and the active force sensor. The two coefficients alpha_f and alpha_x represent the force and displacement scaling factors respectively to adjust the scale between the macro and the micro environments. The force was scaled up by alpha_f=0.5 105. The position handle was scaled downby the gain alpha_x and the result used as a set point for the transducer carrying the sample holder. For a good trade-off between precision and manipulation comfort, the value of 1/alpha_x was 1.4 10-2.

Stability and transparency analysis

Stability : The whole system can be presented as a series connection of sub-systems (blocks). The system is composed of the

haptic interface connected to the human operator and to the slave device (considered as the sum of the transducer of the sample holder and the force sensor) through direct haptic coupling. The slave device is connected to the environment. Assuming that the user will not voluntarily destabilize the system, his/her hand is considered as a passive system [Colgate et al. 93]. The environment is also assumed to be passive as there is no time delay. As the direct coupling itself is passive, the passivity of the whole system is then guaranteed since the series connection of passive systems yield a passive system as detailed previously. As shown in [Richard et al. 99], if a LTI system is passive, the system is stable when coupled to arbitrary passive system which is itself passive. This provides a sufficient condition for the stability of the micro teleoperation system presented in this paper.

Transparency :

To analyze the transparency of the system, the impedance criterion is used in this paper. The operator side impedance Z_op=F_op/X_h

is compared with the environment side impedance Z_e=F_e/X_e. The scaling gains are also taken into account. The perfect transparency

is achieved [Lawrence-1993] if :

Z_op=alpha_f/alpha_d*Z_e

The transparency investigation is carried out for two specific cases: a slave in non-contact mode (no applied force) and a slave in contact mode. For the first mode (non-contact mode), the operator-side impedance depends only on the characteristics of the haptic interface's small

motor. As shown in [Millet-et-al-09], the choice of this motor is made to ensure minimal inertia. Thus, the ideal transparency is achieved for

this mode. For the second mode (contact mode), the environment is modelled as a sample with a stiffness k. As shown in Fig. 4, the Bode diagram of Z_op and of alpha_f/alpha_dZ_e for three values of $k_e$ representing typical stiffness values of soft matter shows that the impedance of the micro environment is transmitted to the human operators for k<1N/m. The bandwidth of the sensor currently limits the stiffness of the samples that can be sensed up to 1N/m. By increasing the sensor bandwidth, the system could interact with stiffer samples.

System validation

The system was experimentally demonstrated in the complex case which consists in measuring the time-course interaction of a thin glass probe with a water droplet under direct human control (see Fig.5 a,b,c). The task is split in four main phases, namely ‘approach’, ‘contact’, ‘retract’, and ‘contact break’ (see Fig.5). The result, Fig.5, shows that the system remains stable over the experiment and achieves a quasi perfect transparency. In fact, the force felt by the user through the haptic interface is exactly the force measured by the sensor amplified by a constant coefficient. Hence, this result shows the performance and the importance of haptic micro teleoperation systems. The performance reached by this new haptic device make it suitable for micro/nanomanipulation where the operator experience is required to achieve complex tasks.

Fig.5 : Interaction of a tiny glass probe with a water droplet. (a) Evolution of the measured force as a function of time and as a function of gap probe-droplet. (b) Force felt by the operator hand during the interaction as a function of time and as a function of handle displacement, that accurately replicates the microscopic interaction. (c) The error between the estimation of felt force and the amplification of the measured force. This error, less then 1%, shows the high degree of transparency of the system, where transmitted force replicates accurately measured microscopic force.

Conclusion

A micro teleoperation system with haptic feedback designed. This system uses an active micro-force sensor, where the measurement was provided by an electrostatic actuator controlled to cancel the action of interaction forces. These forces were transmitted to the operators using a direct force feedback scheme and through a haptic interface capable of very high rendering fidelity. This dual stage device employed two coupled motors for a single-axis force feedback and achieves a ratio of 1000 between the maximum and minimum displayed forces that can be as low as 5mN.

The force felt by the user was a pure homothetic function of the real interaction owing to the transparency of both the sensor and the haptic interface. The overall stability was guaranteed by the passivity of each component. The system was validated by interactively probing the interaction forces of a glass probe with a water droplet over approach-retraction cycles. The system showed great promise in several microscale applications, in particular, the characterization and manipulation of soft matter such as in biological samples, or in the measurement of non-contact forces.