The goal of the work is restore a quadruped walking spider-type robot named Jaromír. Part of the work is a body design that is then printed on a 3D printer. Another part is the design and creation of a new control board, which’ll include sensor equipment appropriate for robot operation. Subsequently, the thesis deals with the design of the tactile sensor’s and their fitting on the legs. The final part is implementation of the algorithm for robot control. The goal was to put the model to a state where it could be used for learning and presentations.

No description available.

In recent times robots are becoming more and more common. They are everywhere. Walking, running, swimming, flying and many of them have much in common with the creatures inhabiting this planet. A lot of it in order to make them appeal more to us, instead of simply being portrayed as stone cold machines. Continuing on the path evolution has laid out before us seems to be a wise decision to make, aspiring to efficiently utilize our knowledge about science and engineering with the vision of improving our future. With the intention to simulate a four legged animal and evaluate the means of interacting with one´s surrounding, a quadruped locomotion system together with two types of sound and voice interacting systems have been assessed. A demonstrator was built to test the real world problems and decide what kind of interacting that is most beneficial. The results indicate that voice commands and speech recognition, rather than sounds from the environment are more practical and robust as a way of interacting with one´s surroundings. / På senare tider har robotar blivit mer och mer vanliga. De är överallt. Gående, springande, simmande, flygande och många av dem har mycket gemensamt med de varelser som lever på denna jord. Mycket av detta för att tilltala oss mer, istället för att framstå som enbart iskalla maskiner. Att fortsätta på den väg som evolutionen har lagt framför oss verkar vara ett vist beslut att ta, i strävan efter att effektivt utnyttja våra kunskaper i vetenskap och ingenjörskonst med visionen om att förbättra vår framtid. Med målet att simulera ett fyrbent djur och utvärdera möjligheterna till att interagera med ens omgivning, har ett fyrbent förflyttningssystem tillsammans med två typer av ljud och röstsystem tagits fram. En prototyp kontruerades för att testa de problem som uppstår i den verkliga värden och för att kunna bedöma vilket sätt att interagera som visar vara sig mest fördelaktigt. Resultaten indikerar att röstkommandon och röstigenkänning, snarare än ljuddetektion från omgivningen är mer praktiska och robusta som ett sätt att interagera med sin närmiljö.

The research presented in this thesis regards walking of quadruped robots, and particularly the walking of the Warp1 robot. The motivation for the robot is to provide a platform for autonomous walking in rough terrain. The thesis contains six papers ranging from development tools to actuation of robot legs. The first paper describes the methods and tools made for control development. These tools feature: programming of the robot without low level coding (C-code); that the controller has to be built only once for simulation and experiments; and that names of variables and constants are unchanged through the chain of software Maple -- Matlab -- Simulink -- Real~Time~Workshop -- xPC--Target. Three controllers, each making the robot walk are presented. The first controller makes the robot walk using the crawl gait. The method uses static stability as method for keeping balance and the instantaneous trunk motions are given by a concept using the so called weight ratios. A method for planning new footholds based on the positions of the existing footholds is also proposed and the controller experimentally verified. The second walking controller shows that the robot also can walk dynamically using the trot gait. The method proposed uses information from ground contact sensors on the feet as input to control balance, instead of, which is common, inertial sensors. It is experimentally verified that Warp1 can trot from level ground onto a slope and turn around while staying balanced. The main ideas of these two walking controllers are fused in the third which enables smooth transitions between crawl and trot. The idea of using the ground contact sensors from the first controller is here used to estimate the position of the center of mass. This controller uses weight ratios in the gait crawl as well as in the dynamic gait trot. Hence, the method of using weight ratios is not only useful for static stability for which it was originally intended. The controller is experimentally verified on Warp1. The Warp1 robot weighs about 60 kg, has 0.6 m long legs with three actuated joints on each. The speed and strength is sufficient only for slow walking, even though the installed power indicates that it should be enough for faster walking. The reason is that a walking robot often needs to be strong but slow when the feet are on the ground and the opposite when in the air. This can not be achieved with the motors and transmissions currently used. A transmission called the passively variable transmission (PVT) is proposed which enhance motor capabilities of robot joints. It is elastic, nonlinear and conservative. Some general properties for elastic transmissions are derived such that they can be compared with conventional transmissions. The PVT gives strong actuation at large loads and fast actuation at small loads. The proposed transmission is compared to a conventional transmission for a specific task, and the result is that a smaller motor can be used. / QC 20100831

Successful development of mechatronic systems requires a combination of targeted hardware and software design. The compact rescue robot (CRR), a quadruped pneumatically-actuated walking robot that seeks to use the benefits garnered from pneumatic power, is a prime example of such a system. This thesis discusses the development and testing of a simulation that will aid in further design and development of the CRR by enabling users to examine the impacts of pneumatic actuation on a walking robot. However, development of an entirely new dynamic simulation specific to the system is not practical. Instead, the simulation combines a MATLAB/Simulink actuator simulation with a readily available C++ dynamics library. This multi-platform approach results in additional incurred challenges due to the transfer of data between the platforms. As a result, the system developed here is designed in the fashion that provides the best balance of realistic behavior, model integrity, and practicality. An analytically derived actuator model is developed using classical fluid circuit modeling together with nonlinear area and pressure curves to model the valve and a Stribeck-Tanh model to characterize the effects of friction on the cylinder. The valve model is designed in Simulink and validated on a single degree-of-freedom test rig. This actuator model is then interfaced with SrLib, a dynamics library that computes dynamics of the robot and interactions with the environment, and validated through comparisons with a CRR prototype. Conclusions are focused on the final composition of the simulation, its performance and limitations, and the benefits it offers to the system as a whole.

Quadruped robots offer distinct advantages in navigating challenging terrains due to their flexible and shock-absorbing characteristics. This flexibility allows them to adapt to uneven surfaces, enhancing their maneuverability. In contrast, rigid robots excel in tasks that require speed and precision but are limited in their ability to navigate complex terrains due to their restricted motion range. Another category of robots, known as soft robots, has gained attention for their unique attributes. Soft robots are characterized by their lightweight and cost-effective design, making them appealing for various applications. Recent advancements have made significant strides in practical control strategies for soft quadruped robots, particularly in diverse and unpredictable environments. An emerging approach in enhancing the autonomy of robots is through reinforcement learning. While this approach shows promise in enabling robots to learn and adapt to their surroundings, it necessitates rigorous training and must exhibit robustness in real-world scenarios. Moreover, a significant hurdle lies in bridging the gap between simulations and reality, as models trained in idealized virtual environments often struggle to perform as expected when deployed in the physical world. This thesis aims to address these challenges by optimizing the control of soft quadruped robots using a model-based reinforcement learning approach. The primary goal is to refine the gait control of these robots, taking into account the complexities encountered in real-world environments. The report covers the implementation of model-based reinforcement learning, including simulation setup, reward design, and policy refinement. Results show improved training efficiency and autonomous behavior, confirming the method’s effectiveness in enhancing soft quadruped robot capabilities.It is important to note that this report provides a concise summary of the thesis results due to the word limit imposed by the Department of Machine Design. For a comprehensive understanding of the research and its implications, the complete version is attached separately here. / Fyrbenta robotar är tack vare deras flexibla och stötdämpande egenskaper är väl lämpade att navigera utmanande terräng. Deras struktur möjliggör anpassning till ojämnheter i underlaget och bidrar till att öka deras rörelseförmåga. I kontrast utmärker sig stela robotar som det bästa valet för uppgifter som kräver snabbhet och precision, men deras förmåga att navigera komplex terräng är begränsad av deras rörelseomfång. En annan typ av robot, så kallade mjuka robotar, har nyligen uppmärksammats för sina unika egenskaper. Dessa robotar kännetecknas av en kostnadseffektiv lättviktsdesign, vilket gör dem attraktiva för många olika användningsområden. Nyligen har betydelsefulla framsteg gjorts inom kontroll av mjuka fyrbenta robotar, framför allt vad gäller kontroll i varierade miljöer. En av de huvudsakliga utmaningarna för att öka robotars autonomi är förstärkningsinlärning. Även om denna teknik är lovande för att ge robotar förmågan att lära sig och anpassa sig efter sin omgivning, kräver den omfattande träning samt måste uppvisa robusthet i verkliga scenarion. Ett större hinder är dessutom klyftan mellan simulation och verklighet, då modeller som tränats i ideella simuleringar ofta presterar sämre än väntat i den fysiska världen. Detta examensarbete behandlar dessa utmaningar genom att implementera en modellbaserad förstärkningsinlärningsmetod för kontroll av fyrbenta robotar, med det primära målet att förfina gångkontrollen för dessa robotar med hänsyn till de komplexa beteenden som uppstår i verkliga miljöer. Denna rapport behandlar implementeringen av modellbaserad förstärkningsin lärning samt simulering, belöningsdesign och policyförfining. Resultat visar på en förbättrad inlärningsförmåga och bättre autonomt beteende, vilket gör metoden lämplig för att förbättra prestandan av mjuka fyrbenta robotar. Var god att notera att denna rapport endast ger en nedkortad sammanfattning av forskningen och dess resultat på grund av krav från institutionen för maskinkonstruktion. En fullständig version innehållande mer detaljer kring studien och dess konsekvenser bifogas här.

Legged robots have the potential to cover terrain not accessible to wheel-based robots and vehicles. This makes them better suited to perform tasks, such as search and rescue, in real-world unstructured environments. Pneumatically-actuated, compliant robots are also more suited than their rigid counterparts to work in real-world unstructured environments with humans where unintentional contact may occur. This thesis seeks to combine the benefits of these two type of robots by implementing design methods to aid in the design choice of a 16 degree of freedom (DoF) compliant, continuum-joint quadruped. This work focuses on the design optimization, especially the definition of design metrics, for this type of robot. The work also includes the construction and closed-loop control of a four-DoF continuum-joint leg used to validate design methods.We define design metrics for legged robot metrics that evaluate their ability to traverse unstructured terrain, carry payloads, find stable footholds, and move in desired directions. These design metrics require a sampling of a legged-robot's complete configuration space. For high-DoF robots, such as the 16-DoF in evaluated in this work, the evaluation of these metrics become intractable with contemporary computing power. Therefore, we present methods that can be used to simplify and approximate these metrics. These approximations have been validated on a simulated four-DoF legged robot where they can tractably be compared against their full counterparts.Using the approximations of the defined metrics, we have performed a multi-objective design optimization to investigate the ten-dimensional design space of a 16-DoF compliant, continuum-joint quadruped. The design variables used include leg link geometry, robot base dimensions, and the leg mount angles. We have used an evolutionary algorithm as our optimization method which converged on a Pareto front of optimal designs. From these set of designs, we are able to identify the trade-offs and design differences between robots that perform well in each of the different design metrics. Because of our approximation of the metrics, we were able to perform this optimization on a supercomputer with 28 cores in less than 40 hours.We have constructed a 1.3 m long continuum-joint leg from one of the resulting quadruped designs of the optimization. We have implemented configuration estimation and control and force control on this leg to evaluate the leg payload capability. Using these controllers, we have conducted an experiment to compare the leg's ability to provide downward force in comparison with its theoretical payload capabilities. We then demonstrated how the torque model used in the calculation of payload capabilities can accurately calculate trends in force output from the leg.

Quadruped robots can traverse a multitude of terrains with greater ease when compared to wheeled robots. Traditional rigid quadruped robots possess severe limitations as they lack structural compliance. Most of the existing soft quadruped robots are tethered and are actuated using pneumatics, which is a low grade energy source and lacks viability for long endurance robots. The work in this thesis proposes the development of a continuum actuator driven quadruped robot which can provide compliance while being un-tethered and electro-mechanically driven. In this work, continuum actuators are developed using mostly 3D printed parts. Additionally, the closed loop control of continuum actuators for walking is developed. Linear Quadratic Regulator (LQR) and pole placement based methods for controller synthesis were evaluated and LQR was determined to be better when minimizing the actuator effort and deviation from set-point. These continuum actuators are composed together to form a quadruped. Gait analyses on the quadruped were conducted and legs of the quadruped were able to trace the gaits for walking and galloping. / Fyrfotarobotar kan lättare korsa en mängd olika terränger jämfört med hjulrobotar. Traditionella styva fyrfotarobotar har kraftiga begränsningar då de saknar strukturell följsamhet. De flesta befintliga mjuka fyrbenta robotar är kopplade till en eller flera kablar och drivs av pneumatik, vilket är en lågkvalitativ energikälla och lämpar sig inte för robotar med lång uthållighet. Arbetet i denna avhandling föreslår utvecklingen av en continuum ställdonsdriven fyrfotarobot, som ger följsamhet samtidigt som den ¨ar frånkopplad och elektromekaniskt driven. I detta arbete framställs continuum ställdon med mestadels 3D-printade delar. Dessutom utvecklas dessa ställdons slutna kontrolloop för gång. Linjärkvadratisk regulator (LQR) och metoder baserade på polplacering utvärderades för styrsyntes, och det fastställdes att LQR presterade bättre när man minimerar ställdonets ansträngning samt avvikelse från referensvärde. Continuum ställdon sammansattes för att bilda en fyrbent robot. Gånganalyser utfördes på roboten och dess ben kunde följa gång- och galopprörelser.

A soft quadruped robot is being developed at the Department of Machine Design and Department of Production Engineering at KTH. The legs of the robot consist of four continuum actuators that can achieve complex movements. In order to efficiently develop gaits for the robot, reinforcement learning will be used. The learning process will use data from simulation instead of directly from the real robot to save time and resources. However, it is significantly more computationally expensive to simulate soft robotics than rigid, because the physical laws of flexible materials are inherently complex. Because of this, soft robot simulations tend to be slower which limits their usability for reinforcement learning. This thesis explores simulation modelling options in Matlab Simscape for the soft quadruped robot, that can be used in reinforcement learning. Four simulation models of the soft actuator were implemented in order to be tested and compared. Two actuation methods and two build options were chosen based on the literature study and related works, and were then permuted for the different combinations. The tested combinations are: lumped-parameter method actuated by internal force, flexible beam actuated by internal force, lumped-parameter method actuated by cable/pulley network and flexible beam actuated by cable/pulley network. The four actuators were built and tested separately. Computational time and simulation-to-reality gap were used for evaluating the modeling methods. The results show that the best option when modelling the soft actuator for reinforcement learning in Matlab Simscape is to use the lumped-parameter method in combination with a cable and pulley network. High accuracy level can still be achieved despite not keeping the true number of attachment points between the cable and actuator. The number of pulleys in the model is linearly correlated to the time cost required to simulate the model. / En mjuk fyrbent robot är under utveckling vid institutionen för maskinkonstruktion och institutionen för industriell produktion på KTH. Robotens ben består av fyra kontinuerligt deformerbara ställdon som kan åstadkomma komplexa rörelser. För att effektivt utveckla gångstilar till roboten kommer förstärkt inlärning att användas. Inlärningsprocessen kommer att använda data från simulering istället för från den fysiska roboten för att spara tid och resurser. Det är dock betydligt dyrare beräkningsmässigt att simulera mjuk robotik än styv, eftersom flexibla material är mer komplexa. På grund av detta tenderar simuleringar av mjuka robotar att vara långsammare, vilket begränsar deras användbarhet för förstärkt inlärning. Detta examensarbete utforskar därför alternativ för modellering och simulering av den mjuka fyrbenta roboten i Matlab Simscape, med målet att den ska kunna användas med förstärkt inlärning. Fyra olika simuleringsmodeller av det mjuka ställdonet implementerades för att testas och jämföras. Två aktiveringsmetoder och två konstruktionsalternativ valdes baserat på litteraturstudien och relaterade arbeten, och permuterades sedan till möjliga versioner. De testade versionerna är således: klumpparametermetod som aktiveras av intern kraft, flexibel balk som aktiveras av intern kraft, klumpparametermetod som aktiveras av kabelnätverk och flexibel balk som aktiveras av kabelnätverk. De fyra ställdonen byggdes och testades separat. Beräkningstid och grad av verklighetstrogenhet, användes för att jämföra resultaten av dessa tester. Resultaten visar att det bästa alternativet vid modellering av det mjuka ställdonet för förstärkt inlärning i Matlab Simscape är att använda klumpparametermetoden i kombination med ett kabelnätverk. Hög noggrannhetsnivå kan uppnås trots att man inte bibehåller det verkliga antalet fästpunkter mellan kabeln och ställdonet. Antalet fästpunkter för kabeln i modellen är linjärt korrelerat till den tidskostnad som krävs för att simulera modellen.