IROS 2015 Papers

This paper presents a systematic flight controller design based on the mathematics of parametrized manifolds and calculus of variations for the Bat Bot (B2), which possesses many articulated wings. Wing kinematics and morphological properties are crucial in the powered flight of flying vertebrates. The articulated skeleton of these mammals, which contains many degrees of actuation and underactuation, has made it difficult to understand the connection between the bat's flight dynamics and its intricate array of physiological and morphological specializations. B2 is a biomimetic micro aerial vehicle (MAV) that possesses similar morphological properties to a bat in order to duplicate bats powered ballistic motion. In an effort to design the advanced flight control algorithm for B2, this paper reports two major contributions. First, a systematic mathematical framework is introduced that evaluates the holonomically-constrained Lagrangian model of a flapping robot with specified active and passive degrees of freedom (DoF) in order to locate physically feasible and biologically meaningful periodic solutions using optimization. These are parametrized constraint manifolds; the flapping wing dynamics are governed by these manifolds. Second, calculus of variations and the well-recognized method of inverse dynamics are applied in order to synthesize the flight control algorithm for the flapping wings.

State-of-the-art techniques for simultaneous localization and mapping (SLAM) employ iterative nonlinear optimization methods to compute an estimate for robot poses. While these techniques often work well in practice, they do not provide guarantees on the quality of the estimate. This paper shows that Lagrangian duality is a powerful tool to assess the quality of a given candidate solution. Our contribution is threefold. First, we discuss a revised formulation of the SLAM inference problem. We show that this formulation is probabilistically grounded and has the advantage of leading to an optimization problem with quadratic objective. The second contribution is the derivation of the corresponding Lagrangian dual problem. The SLAM dual problem is a (convex) semidefinite program, which can be solved reliably and globally by off-the-shelf solvers. The third contribution is to discuss the relation between the original SLAM problem and its dual. We show that from the dual problem, one can evaluate the quality (i.e., the suboptimality gap) of a candidate SLAM solution, and ultimately provide a certificate of optimality. Moreover, when the duality gap is zero, one can compute a guaranteed optimal SLAM solution from the dual problem, circumventing non-convex optimization. We present extensive (real and simulated) experiments supporting our claims and discuss practical relevance and open problems.

This paper addresses one of the fundamental tasks for the aerial assembly of tensile structures: aerial knot-tying. It presents a framework for representing and realizing knots with flying machines. A suitable representation of the knot topology is introduced taking into account the use of supporting elements and the characteristics of flying machines. This information is then translated into three-dimensional trajectories for the vehicle performing the aerial knot-tying task. Furthermore, preliminary results suggest that the quality of the resulting knot can be improved by the use of an iterative learning algorithm. Experiments are performed with quadrocopters to validate the proposed approach. An accompanying video shows the aerial knot-tying process.

Analysis of human running has revealed that the motion of the human leg can be modeled by a compression spring because the leg's joints behave like a torsion spring in the stance phase. Moreover, the knee bends rapidly to avoid contact of the foot with the ground in the swing phase. In this paper, we describe the development of a knee joint mechanism that mimics the elastic characteristics of the stance leg and rapid bending knee of the idling leg of a running human. The knee was equipped with a mechanism comprising two leaf springs and a worm gear for adjusting the joint stiffness and high-speed bending knee. Using this mechanism, we were able to achieve joint stiffness within the range of human knee joints that could be adjusted by varying the effective length of one of the leaf springs. In addition, the mechanism was able to bend rapidly by changing the angle between the two leaf springs. The equation proposed for calculating the joint stiffness considers the difference between the position of the fixed point of the leaf spring and the position of the rotational center of the joint. We evaluated the performance of the adjustable joint stiffness and the effectiveness of the proposed equation for joint stiffness and high-speed knee bending. We were able to make a bipedal robot hop using pelvic oscillation for storing energy produced by the resonance to leg elasticity and confirmed the mechanism could produce large torque 210 Nm.

Much attention has recently been given to a printing method because they are easily designable, have a low cost, and can be mass produced. Numerous electronic devices are fabricated using printing methods because of these advantages. In paper mechatronics, attempts have been made to fabricate robots by printing on paper substrates. The robots are given structures through self-folding and functions using printed actuators. We developed a new system and device to fabricate more sophisticated printed robots. First, we successfully fabricated complex self-folding structures by applying an automatic cutting. Second, a rapidly created and low-voltage electrothermal actuator was developed using an inkjet printed circuit. Finally, a printed robot was fabricated by combining two techniques and two types of paper; a structure design paper and a circuit design paper. Gripper and conveyor robots were fabricated, and their functions were verified. These works demonstrate the possibility of paper mechatronics for rapid and low-cost prototyping as well as of printed robots.

The Space Exploration Guided Heuristic Search (SEHS) method solves the motion planning problem, especially for car-like robots, in two steps: a circle-based space exploration in the workspace followed by a circle-guided heuristic search in the configuration space. This paper extends this approach for kinodynamic planning in dynamic environments by performing the exploration in both space and time domains. Thus, a time-dependent heuristic is constructed to guide the search algorithm applying a kinodynamic vehicle model. Furthermore, the search step-size and state resolution are adapted incrementally to guarantee resolution completeness with a trade-off for efficiency. The performance of Space-Time Exploration Guided Heuristic Search (STEHS) approach is verified in two scenarios and compared with several search-based and sampling-based methods.

This paper presents kinematics and two dynamic characteristics including stiffness and natural frequency of a four degrees-of-freedom (DOFs) masticatory robot. Based on the biological architecture of human masticatory system, this robot is modeled in the form of a redundantly actuated parallel mechanism constrained by two point-contact higher kinematic pairs (HKPs). The aim of this robot is to chew samples of newly developed foods in a human way while the food textures can be characterized. The kinematics including inverse kinematics, workspace and singularity are studied systematically. The inverse kinematics is given in closed-form solutions. The workspace is generated by a numerical search method. In virtue of the Jacobian matrix derived from differential kinematics, the singularity and the stiffness are investigated. Via the derived stiffness matrix and mass matrix, natural frequency is also computed. Finally, simulation is carried out, one real incisor trajectory of a healthy human subject is reproduced, illustrating the robot is able to reproduce kinematic behaviors of the human masticatory system, and it owns a satisfactory performance in terms of stiffness and natural frequency. Comparison between the redundantly actuated masticatory robot and its two non-redundantly actuated counterparts also exhibits that the former owns better performance, in terms of a similar incisor's workspace as the Posselt envelope, no singularity, larger stiffness and higher natural frequency.

This paper challenges the well-established assumption in robotics that in order to control a robot it is necessary to know its kinematic information, that is, the arrangement of links and joints, the link dimensions and the joint positions. We propose a kinematic-free robot control concept that does not require any prior kinematic knowledge. The concept is based on our hypothesis that it is possible to control a robot without explicitly measuring its joint angles, by measuring instead the effects of the actuation on its end-effector. We implement a proof-of-concept encoderless robot controller and apply it for the position control of a physical 2-DOF planar robot arm. The prototype controller is able to successfully control the robot to reach a reference position, as well as to track a continuous reference trajectory. Notably, we demonstrate how this novel controller can cope with something that traditional control approaches fail to do: adapt to drastic kinematic changes such as 100% elongation of a link, 35-degree angular offset of a joint, and even a complete overhaul of the kinematics involving the addition of new joints and links.

The use of more than one steerable (standard) driving wheel allows a robot to perform omnidirectional motions. However, the modeling and control of such robots is challenging since the system is non-holonomic, nonlinear and typically over actuated. Moreover, such platforms exhibit kinematic singularities. A well known singular configuration is the configuration where two steerable driving wheels are coaxial aligned. This is highly problematic since this configuration corresponds to pure rotations, which is crucial for narrow space navigation. In this paper a control scheme with improved robustness w.r.t. these singularities is derived. It is based on the second order (accelerations) non-holonomic constraints. The remaining singularity is tackled by a regular parametrization of the robot's motion. Thereupon a novel control concept is presented which is based on an input-output linearization in terms of a path parameter. The choice of this parametrization provides an additional parameter in the controller design. The approach is demonstrated for a prototype implementation.

An algorithm is proposed for matching data from different sensing modalities. The problem is formalised as a kidnapped robot problem, where Bayesian fusion is used to find the most likely location where both modalities agree. The key idea of our algorithm is to model the correlation between the two modalities as a likelihood used to update a location prior. Data, in this case, is represented as 2.5D thickness maps from a laser scanner and a Remote Field Eddy Current (RFEC) tool, used in non-destructive testing to assess the condition of infrastructures. The laser data is limited, while RFEC data is continuous. Given some prior in location, the aim is to find the 2.5D thickness map from the laser that corresponds to the RFEC data, which should be noted is highly noisy. Real data from CCTV inspections of water pipes are used to validate the proposed approach.

This paper proposes the use of a kernel density estimation to measure similarities between trajectories. The similarities are then used to predict the future locations of a target. For a given environment with a history of previous target trajectories, the goal is to establish a probabilistic framework to predict the future trajectory of currently observed targets based on their recent moves. Instead of clustering trajectories into groups, we calculate the similarity between a given test trajectory and the set of all past trajectories in a dataset. Next, we use a weighted mechanism for prediction, that can be used in target tracking and collision avoidance applications. The proposed method is compared with two other commonly used similarity models (PCA and LCSS) over a dataset of simulated trajectories, and two datasets of real observations. Results show that the proposed method significantly outperforms the existing models for those datasets and experimental settings.

Robust, scalable localization unlocks path-planning, obstacle avoidance as well as manipulation and thus is a core competency for many robotic applications. However, as we leave the lab and move out in the world, models of the environment no longer span distances of meters but kilometers in length. Now, gigabytes instead of megabytes of memory are required to hold the model of the environment required for localization. Discarding data and keeping the map representation compact is thus essential for any meaningful application. This paper presents and evaluates a map compression algorithm that approaches this data-reduction as an constrained optimization problem. At the core of the algorithm is the concept of a Summary Map, a reduced map representation that includes only the landmarks that are deemed most useful for place recognition. To assign landmarks to the map we have to satisfy the conflicting goals of map coverage and localizability as well as our tight memory budget. While using an optimization approach for compression is not novel, in this paper we propose adaptations to drastically reduce the computational requirements. Our approach improves scalability from trajectories of a few tens of meters manageable by the state of the art to virtually unlimited dataset sizes in our system. We evaluate the performance of various compression levels as well as several methods for selecting the best localization landmarks from outdoor datasets.

In general, the problems of objects' and rooms' categorizations for robotic applications have been addressed separately. The current trend is, however, towards a joint modelling of both issues in order to leverage their mutual contextual relations: object → room (e.g. the detection of a microwave indicates that the room is likely to be a kitchen), and room → object (e.g. if the robot is in a bathroom, it is probable to find a toilet). Probabilistic Graphical Models (PGMs) are typically employed to conveniently cope with such relations, relying on inference processes to hypothesize about objects' and rooms' categories. In this work we present a Conditional Random Field (CRF) model, a particular type of PGM, to jointly categorize objects and rooms from RGBD images exploiting object-object and object-room relations. The learning phase of the proposed CRF uses Human Knowledge (HK) to eliminate the necessity of gathering real training data. Concretely, HK is acquired through elicitation and codified into an ontology, which is exploited to effortless generate an arbitrary number of representative synthetic samples for training. The performance of the proposed CRF model has been assessed using the NYU2 dataset, achieving a success of ~ 70% categorizing both, objects and rooms.

This paper proposes an iterative learning control scheme for whole-arm object manipulation by using a robotic hand system composed of a velocity-controlled finger and a torque-controlled finger. The novel iterative learning control scheme consists of the PD-type iterative learning control scheme for the velocity-controlled finger to follow the desired trajectory and the optimization-based iterative learning control scheme for the torque-controlled finger to maintain equilibrium. The switching between the two controllers is made to achieve compliant manipulation. The validity of the proposed scheme is demonstrated through experiments

In this paper, we address the problem of gait generation for bipedal robots. We cast the determination of a Center of Mass (CoM) reference trajectory for a given Zero Moment Point (ZMP) desired behaviour as a stable inversion problem for non-minimum phase systems and obtain an analytical solution for any given ZMP trajectory. Our method exploits results from our previous research, in which we derived a family of bounded CoM trajectories associated to a given desired ZMP trajectory.

Search and rescue (SAR) missions are being carried out by several types of robots. They include ground, marine and air vehicles depending on the terrain and mission to be tackled. A particular niche for SAR activities are shallow waters. They present high difficultly for conventional ground or marine robots because of the mix of water and ground. Such an environment is difficult to be accessed for a robot without some built-in amphibious capabilities. Our lab has experience in the design of amphibious salamander-like robots. In order to consider whether these robots would be suited for SAR missions in shallow waters, a key requirement is the ability to tackle rough terrains. In this paper we present a control framework for a highly redundant salamander-like robot. It involves bio-inspired spine control, inverse kinematics-based limb control, proper limb-spine coordination, reflex mechanisms and attitude control. The framework is validated in a simulation and on the real robot. In both cases, the robot is used in two different configurations: with and without its tail, in order to investigate how the tail (which is necessary for swimming) affects ground locomotion. With this exploration, we aim to set the precedent for improving the problem of dynamic locomotion of salamander-like robots over unperceived rough terrain. Our results confirm that the design of reflexes like stumbling and extension, combined with an attitude controller, allows for the improving of the performance of the robot in a generic rough terrain which includes stairs, holes and bumps with several levels of complexity adjusted according to the robot dimensions.

Advances in miniature surgical instrumentation are key to less invasive and safer medical interventions. In cardiovascular procedures interventionalists turn towards catheter-based interventions, treating patients considered unfit for classical more invasive approaches. Improvements in design and steerability of catheters could further reduce the invasiveness of these interventions. For example, by improving controllability and interaction forces with the vessels, tissue damage could be limited. Through improved steerability and coordinated control, operation times and exposure to radiation might also be reduced. Latter argument formed the original motivation for the development of teleoperated robotic catheters. Despite the large kinematic dissimilarity and thus non-trivial mapping between joystick input and catheter output motion, few investigations have been conducted to find intuitive mappings that allow straightforward catheter steering. This paper presents some recent work in this direction. Three promising mappings are proposed. The mappings were implemented and validated upon a robotic catheter moving inside an artificial aorta model. Experimental results show good steerability of the robotic catheter for all the mappings. Although superiority of one mapping with respect to the others was observed, further investigation and validation is planned. In the future, additional visual cues that increase the situational awareness of the user are expected to further simplify the steering.

Assistive devices such as exoskeletons are capable of providing rehabilitative improvement and independence for individuals suffering from musculoskeletal conditions. Typical devices use either active assistance methods such as DC motors or passive methods such as springs. Active methods require a continuous power input, while passive methods are limited by user capability. This work introduces an Active/Passive EXoskeleton (APEX) framework. This device can passively provide continuous assistance, only requiring energy to change the dynamic properties of the passive state. The first prototype (APEX-α) is introduced and tested on six healthy subjects who performed hammer curls. It was found that changes in the passive state of the APEX-α affect the number of curls performed by an individual. By changing the passive state of the exoskeleton, increases in curl count of 65 - 92% were observed. This indicates the potential for such devices to provide assistance to an individual through the use of lightweight, energy efficient active/passive actuators.

We present MantisBot, a 28 degree of freedom robot controlled by a high-fidelity neural simulation. It is modeled after the mantis, with many degrees of freedom, because we intend to study directed behaviors and leg multi-functionality, such as prey tracking and striking. As a first step, we present a distributed reflexive posture controller. MantisBot maintains posture through a series of reflexes observed in insects, specifically: strain measurements from a leg produce proportional torque commands (reflex A); large or rapidly decreasing leg strains produce a rapid, single “restep” (reflex B); a leg can only restep if its neighboring legs are all under strain (reflex C); and a leg will search for the ground if it does not reach it as expected (reflex D). All of these reflexes contribute to a hardware platform's posture, and are implemented in a highly distributed fashion. The two most distal joints in each leg each has its own central pattern generator (CPG, 12 total), upon which all of these behaviors depend. To achieve the desired dynamics, we implement a control network of conductance-based neurons with persistent sodium channels arranged in a network like the animal may possess in its thoracic ganglia. The result is a robot capable of actively maintaining posture without a centralized planner or body model. In addition, the network implementation is fast, calculating network dynamics 150 times faster than real time.