An official website of the United States government
Abstract Inspired by animals that co-adapt their brain and body to interact with the environment, we present a tendon-driven and over-actuated (i.e.njoint,n+1 actuators) bipedal robot that (i) exploits its backdrivable mechanical properties to manage body-environment interactions without explicit control,and(ii) uses a simple 3-layer neural network to learn to walk after only 2 min of ‘natural’ motor babbling (i.e. an exploration strategy that is compatible with leg and task dynamics; akin to childsplay). This brain–body collaboration first learns to produce feet cyclical movements ‘in air’ and, without further tuning, can produce locomotion when the biped is lowered to be in slight contact with the ground. In contrast, training with 2 min of ‘naïve’ motor babbling (i.e. an exploration strategy that ignores leg task dynamics), does not produce consistent cyclical movements ‘in air’, and produces erratic movements and no locomotion when in slight contact with the ground. When further lowering the biped and making the desired leg trajectories reach 1 cm below ground (causing the desired-vs-obtained trajectories error to be unavoidable), cyclical movements based on either natural or naïve babbling presented almost equally persistent trends, and locomotion emerged with naïve babbling. Therefore, we show how continual learning of walking in unforeseen circumstances can be driven by continual physical adaptation rooted in the backdrivable properties of the plant and enhanced by exploration strategies that exploit plant dynamics. Our studies also demonstrate that the bio-inspired co-design and co-adaptations of limbs and control strategies can produce locomotion without explicit control of trajectory errors.
more »
« less
This content will become publicly available on August 30, 2026
Dynamic wind-up locomotion enabled by embodied intelligence
Repetitive subtasks of locomotion are offloaded from a conventional computer-actuator-sensor set-up to automatic mechanical processes. The subtasks considered are: (1) when out-of-contact with the environment, move a leg to a ready position in preparation for step contact, and (2) when contact is detected, push off the ground. Using conventional closed-loop control, subtask (1) would be accomplished by programming logic and a feedback loop onto a computer-motor-encoder system, and subtask (2) would be accomplished by sensing contact, then commanding the leg motor to push-off via programmed computer logic. We demonstrate how to transition this programmed logic from a computer processor to a mechanical processor. The mechanical processor performs preprogrammed actions based on combinations of states of components, some of which are internal and some that interact with the environment. Because signals are not digital, but rather mechanical quantities of energy, position, and force; transitioning to a mechanical processor enables a third subtask not possible by the computer alone: that is, (3) the accumulation of elastic energy while out-of-contact with the environment, and its automatic release upon contact for a more powerful push-off motion. Migrating processing out of the computer reduces the number of transduction steps, allows for faster responses to dynamic events, and instantiates a high-powered reflex triggered by ground contact. To illustrate these benefits, a robot with built-in onboard mechanical processing is compared to a conventional robot with logic executed by an offboard computer.
more »
« less
- Award ID(s):
- 2144732
- PAR ID:
- 10656470
- Publisher / Repository:
- Sage Journals
- Date Published:
- Journal Name:
- The International Journal of Robotics Research
- ISSN:
- 0278-3649
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Robotic-assisted minimally invasive surgery (MIS) has enabled procedures with increased precision and dexterity, but surgical robots are still open loop and require surgeons to work with a tele-operation console providing only limited visual feedback. In this setting, mechanical failures, software faults, or human errors might lead to adverse events resulting in patient complications or fatalities. We argue that impending adverse events could be detected and mitigated by applying context-specific safety constraints on the motions of the robot. We present a context-aware safety monitoring system which segments a surgical task into subtasks using kinematics data and monitors safety constraints specific to each subtask. To test our hypothesis about context specificity of safety constraints, we analyze recorded demonstrations of dry-lab surgical tasks collected from the JIGSAWS database as well as from experiments we conducted on a Raven II surgical robot. Analysis of the trajectory data shows that each subtask of a given surgical procedure has consistent safety constraints across multiple demonstrations by different subjects. Our preliminary results show that violations of these safety constraints lead to unsafe events, and there is often sufficient time between the constraint violation and the safety-critical event to allow for a corrective action.more » « less
-
This paper describes the Stevens Institute of Technology's submission for the WMT 2022 Shared Task: Code-mixed Machine Translation (MixMT). The task consisted of two subtasks, subtask 1 Hindi/English to Hinglish and subtask 2 Hinglish to English translation. Our findings lie in the improvements made through the use of large pre-trained multilingual NMT models and in-domain datasets, as well as back-translation and ensemble techniques. The translation output is automatically evaluated against the reference translations using ROUGE-L and WER. Our system achieves the 1st position on subtask 2 according to ROUGE-L, WER, and human evaluation, 1st position on subtask 1 according to WER and human evaluation, and 3rd position on subtask 1 with respect to ROUGE-L metric.more » « less
-
The robotic spine has a lot of potential for snake-like, quadruped, and humanoid robots, as it can improve their mobility, flexibility, and overall function. A common approach to developing an articulated spine uses geared motors to imitate vertebrae. Instead of using geared motors that rotate 360 degree, a bioinspired gearless electromechanical actuator was proposed and developed as an alternative, specifically for humanoid spine applications. The actuator trades off angular flexibility for torque, while the geared motor trades off speed for torque. This article compares the proposed actuator and conventional geared motors regarding torque, acceleration, and copper loss for a vertebra's angular flexibility. When its angular flexibility is lower than 14∘, the proposed actuator achieves higher torque capability without gears than with conventional motors. Lower angular flexibility, which means smaller airgaps, allows the proposed actuator to produce a much stronger torque for the same input power. The actuator's nonlinear electrical and mechanical dynamics models are developed and used for position control of a six-module distributed spine. In addition, two different position-control architectures are developed: an outer loop proportional-integral (PI) position controller with an inner loop PI current controller and an outer loop PI position controller with an inner loop PI torque controller.more » « less
-
Jiang, Jingang (Ed.)Most current powered transfemoral prostheses are designed based on replicating normal anatomy with the inclusion of a revolute knee joint. Prosthesis users often have issues achieving proper leg length to maintain balance and perform push-off during stance, and to ensure sufficient toe clearance during swing. There is a clinical opportunity to develop a powered prosthesis that linearly shortens and lengthens during ambulation with a prismatic joint for improved leg length properties. To build on previous work, the research in this manuscript focuses on designing the physical device, the leg length actuation profile, and the control scheme. Based on gait analyses of two prosthesis users, the device provides an appropriate leg length actuation profile with sufficient shortening for toe clearance (exhibited by the greater prosthetic vs. intact side toe clearance) and lengthening for forward propulsion (exhibited by the ground reaction force peak in late stance). The device also has a motor torque and velocity capable of supporting up to a 90 kg user during normal ambulation, a control scheme with an adjustable actuation cycle based on gait cadence (matching within 2 ms), and a more compact mechanical system design (4.5 kg) less than anatomical weight requirements (5.5 kg). Additionally, the prosthesis users tested were highly encouraging of their stability, mobility, and safety while ambulating with the device.more » « less
