Posture Control-Human-Inspired Approaches for Humanoid Robot Benchmarking: Conceptualizing Tests, Protocols and Analyses

Abstract

Posture control is indispensable for both humans and humanoid robots, which becomes especially evident when performing sensorimotor tasks such as moving on compliant terrain or interacting with the environment. Posture control is therefore targeted in recent proposals of robot benchmarking in order to advance their development. This Methods article suggests corresponding robot tests of standing balance, drawing inspirations from the human sensorimotor system and presenting examples from robot experiments. To account for a considerable technical and algorithmic diversity among robots, we focus in our tests on basic posture control mechanisms, which provide humans with an impressive postural versatility and robustness. Specifically, we focus on the mechanically challenging balancing of the whole body above the feet in the sagittal plane around the ankle joints in concert with the upper body balancing around the hip joints. The suggested tests target three key issues of human balancing, which appear equally relevant for humanoid bipeds: (1) four basic physical disturbances (support surface (SS) tilt and translation, field and contact forces) may affect the balancing in any given degree of freedom (DoF). Targeting these disturbances allows us to abstract from the manifold of possible behavioral tasks. (2) Posture control interacts in a conflict-free way with the control of voluntary movements for undisturbed movement execution, both with "reactive" balancing of external disturbances and "proactive" balancing of self-produced disturbances from the voluntary movements. Our proposals therefore target both types of disturbances and their superposition. (3) Relevant for both versatility and robustness of the control, linkages between the posture control mechanisms across DoFs provide their functional cooperation and coordination at will and on functional demands. The suggested tests therefore include ankle-hip coordination. Suggested benchmarking criteria build on the evoked sway magnitude, normalized to robot weight and Center of mass (COM) height, in relation to reference ranges that remain to be established. The references may include human likeness features. The proposed benchmarking concept may in principle also be applied to wearable robots, where a human user may command movements, but may not be aware of the additionally required postural control, which then needs to be implemented into the robot.

Keywords: benchmarking; human-like versatility and robustness; humanoid robots; posture control; sensorimotor system.

Figure 1

Single inverted pendulum (SIP) scenario of posture control in the body sagittal plane. Shown are the four basic disturbances (on the left) with impact on the ankle joints in terms of disturbing torques, as indicated by the arrows on the right side. T_A, total ankle torque; T_act, commanded active torque; T_ext, external torque from contact force; T_grav from field force gravity; T_in from inertial effect upon foot in space translational acceleration (Ẍ_FS; T_pass, from passive stiffness of muscles and connective tissues.

Figure 2

(A–F) Suggested posture control disturbance scenarios (inspired by human studies). Examples refer to SIP scenarios that challenge balancing of biped standing in the sagittal body plane (with moderate stimuli mainly around the ankle joints). (A) Support surface (SS) rotation about the ankle joint axis (using a hexapod platform; Mergner et al., 2003). (B) Support translation. (C) Contact force stimulus (applied as pull on a body harness using cable winches). (D) “Body sway referencing of the platform” (BSRP). Spontaneous or evoked body sway is recorded and sway signal is used to tilt the support along with the body such that the ankle joint angle (and its proprioception) remains fixed and compensation of the field force gravity with eyes closed requires vestibular input. The effect of visual self-motion and spatial orientation cues are evaluated by comparing in scenarios (A–D) the balancing in “eyes open” and “eyes closed” conditions. (E) Isolated visual scene motions, to test how successful the postural control system can suppress visually-evoked self-motion illusions (given the robot involves visual motion and orientation cues in its postural control; see text). (F) Combinations of two or more disturbances and of superimposing voluntary movements on external disturbances to test conflict-free interaction between proactive and reactive balancing.

Figure 3

Proof of principle examples from robot experiments (Posturob II) for the disturbance scenarios (A–D) and (F) in Figure 2 (see also Table 1). (A) Center of mass (COM) sway responses to pseudo-random ternary sequence stimulus (PRTS) SS tilt of peak-to-peak (pp) 4° (six successive PRTS cycles). (B) Responses to horizontal SS translations (otherwise as in A). (C) COM sway responses to sinusoidal pull stimuli. (D) Spontaneous COM sways with “body-sway referenced platform” (BSRP) and response to a manual push perturbation. (E) Pull responses as in (C) but superimposed on BSRP (no additional push stimulus). (F) Commanded (“voluntary”) lean of leg segment in space (LS) around ankle joints and return to starting position with “raised cosine velocity” (RC) profiles. Associated is, as an emerging property of the DEC control, a counter-lean of trunk in space (TS) in the hip joints towards upright (dashed line), which reduces the COM excursion (TS command was to maintain trunk orientation in hip joints upright).

Figure 4

Frequency response functions (FRF) of sagittal COM sway responses for peak-peak 1° and peak-peak 4° PRTS stimuli of a human subject (A) and Posturob II (B). Gain, phase and coherence functions across frequency characterize the tilt-evoked sway. Gain gives the amplitude ratio between sway response amplitudes and tilt stimulus amplitudes, with a gain of unity indicating that the sway response amplitude equals the stimulus sway, while a gain of zero indicates that the stimulus did not evoke any sway. Phase characterizes the temporal relation between tilt stimulus and sway response. Coherence is a measure of the signal to noise ratio of the stimulus evoked sway. Responses to this stimulus lend themselves to evaluation of human-likeness (see text).

Figure 5

Commanded (“voluntary”) sinusoidal movements of TS of pp 3° at 0.1 Hz while standing on SS maintained level (A) and on SS tilting with PRTS profile (B). The trunk lean movements are associated with counter leans of the LS and less so of the body COM. Command for the ankle joints was to maintain the body COM vertical.

Figure 6

Sway responses of body COM of Posturob II (A) and of Lucy (B) to sinusoidal tilts of the SS at 0.2 Hz around the ankle joints. Note that COM response amplitudes are similar in the two experiments despite considerable differences in the robots’ anthropometrics (see text). Lower case letters in (A) refer to the two degree of freedom (DoF) Posturob II’s human-inspired anthropometrics, actuation, and sensors, and to the hexapod platform in the posture control laboratory: (a) artificial vestibular sensor, see Mergner et al. (2009); (b and e) hip joints with angle/angular velocity sensors; c, pneumatic muscle and f, force sensors for actuation control; (d,e) ankle joints with angle/angular velocity sensors; (g) ground reaction force sensors under heels and forefeet; (h) hexapod platform for tilt, translation, and BSRP). In the 14 DoF Lucy Posturob (B), force-controlled actuation is using spindle drives; other technical features are analogous to those in Posturob II.