In robotics, software intelligence has advanced faster than hardware capability. Artificial intelligence can plan, reason, and simulate, but physical execution still depends on motor–gearbox actuators designed decades ago. This imbalance creates a bottleneck: robots think faster than they move.

Adaptive Machines was founded to remove that bottleneck. We develop actuators with Physical Intelligence—hardware that collaborates with physics rather than fighting against it. Our mission is to make robots safer, faster, stronger, and more energy-efficient by embedding adaptability directly into their mechanical core.

The Actuator Bottleneck

Robotic performance today is constrained by hardware not designed for dynamic, variable, or human-centered tasks. Traditional actuators rely on rigid components that waste energy, add inertia, and reduce responsiveness. Each improvement in artificial intelligence or control software eventually collides with these physical limits.

Nature demonstrates a different approach. Muscles and tendons store and release energy and adjust stiffness to balance speed and safety. Adaptive Machines applies these same principles to engineered systems, creating actuators that interact constructively with physics rather than resisting it.

Why This Matters

Robotics is entering a new era. Collaborative and humanoid robots are moving from prototypes to real deployments. Artificial intelligence is no longer the main barrier; execution and hardware are.

Custom-built actuators have shown that improved performance is possible, but they are expensive and platform-specific. Scaling robotics requires standardized adaptive actuators—off-the-shelf components that combine high performance with broad reusability. Adaptive Machines delivers that foundation.

From Research to Scalable Solutions

Adaptive Machines is built on a decade of research in mechanically adaptive robotics and energy-efficient actuation. This work has demonstrated that hardware can embody intelligence—reducing power consumption and improving interaction safety without sacrificing precision.

We transform these insights into deployable technology, bridging the gap between theory, innovation, and industrial application.

The post Adaptive Machines: The Hardware Advantage appeared first on David Braun.

Robotics is reaching new frontiers, with innovative designs reshaping how machines and humans interact. Mechanically adaptive robotics offers a path to enhancing mobility without relying on external power sources like motors or pneumatic actuators. By focusing on energy-efficient systems inspired by mechanisms like bicycles, this approach integrates programmable springs—mechanical elements with controllable stiffness—to redefine safety, adaptability, and efficiency in both robotic and human applications.

Programmable Springs: A Breakthrough in Mechanical Design

At the core of this research is the concept of programmable springs. These mechanical components allow precise energy storage and release independently of deformation, making them adaptable to various tasks at near-zero energy cost. This innovation eliminates the need for external power while enhancing the capabilities of robots and human-assistive devices. It also enables novel approaches to energy use, such as leveraging resonance-based control to optimize motion and efficiency.

Key Applications of Mechanically Adaptive Robotics

Mechanically adaptive robots have wide-ranging applications, including:

• Assistive Devices for the Elderly: Programmable springs can provide lightweight and efficient mobility assistance, helping older adults move more independently and safely.
• Emergency Response Tools: Robots equipped with these systems can assist first responders in carrying loads, navigating rough terrain, or performing physically demanding tasks.
• Energy-Efficient Transportation: Legged robots with programmable springs could provide sustainable alternatives to bicycles, offering new mobility options in diverse environments.

These examples showcase the potential to create practical, impactful solutions for everyday challenges.

Broader Impacts

This research extends beyond developing better robots—it contributes to advancing mobility, education, and public understanding:

• Redefining Mobility: By combining performance and energy efficiency, mechanically adaptive robotics addresses key challenges in human augmentation and autonomous systems.
• Inspiring Education: The project promotes engineering concepts centered on controlled energy storage and release, encouraging new approaches to mechanical design.
• Engaging the Public: Outreach efforts, including demonstrations and knowledge-sharing, aim to spark curiosity and innovation in robotics and mobility technologies.

Looking Ahead

The findings from this research pave the way for safer, more efficient mobility solutions that address critical needs in healthcare, emergency response, and sustainable transportation. By emphasizing energy efficiency and mechanical adaptability, these systems challenge traditional paradigms in robotics, inspiring innovations that balance practicality and impact.

This work is supported by the National Science Foundation CAREER Award (Award Number 2144551) under the Foundational Robotics Research Program within the Division of Civil, Mechanical, and Manufacturing Innovation.

If this topic resonates with you, I would love to hear your thoughts or discuss potential collaborations. Feel free to reach out and share your perspective.

The post Mechanically Adaptive Robotics: National Science Foundation CAREER Award appeared first on David Braun.

Compliant actuation, inspired by biological systems, offers robots mechanical adaptability and energy efficiency, making them suitable for tasks requiring both precision and power. However, controlling robots with compliant actuators is inherently challenging due to the complexity of their dynamics. My research tackled this problem by developing a constrained optimal variable impedance control framework to enable efficient and adaptive robot motion.

Key Contributions to Robotics

• Optimal Variable Impedance Control Framework:
  This framework optimizes robot performance under actuation constraints, balancing precision, energy efficiency, and adaptability. It provides a structured approach for achieving dynamic tasks with compliant actuators.
• Dynamic Motion Demonstration:
  The framework was applied to the DLR David robot, a highly advanced variable impedance system, in a dynamic throwing task akin to javelin throwing. Precise timing and energy-efficient execution were achieved, resulting in human-like throwing motions generated purely through optimization-based control methods, without relying on human data or demonstrations.

This was possible because, despite appearing as an open-loop control on the level of the robot joints, the optimal control formulation leveraged the intrinsic feedback capabilities of compliant actuators to reliably execute highly dynamic motions and achieve significant performance gains.

Recognition from the Field

This body of work was recognized with the 2014 IEEE Transactions on Robotics King-Sun Fu Memorial Best Paper Award, presented annually to the best paper published in IEEE Transactions on Robotics, a leading journal in the field. The award honors contributions that demonstrate technical merit, originality, practical significance, and potential impact on robotics research and applications.

Looking Forward

The insights from this research provide a foundation for addressing challenges in robotics, including dynamic motion control, energy efficiency, and adaptive performance. These principles have broad applications in areas such as assistive robotics, legged locomotion, and industrial automation.

If you are interested in discussing this work further or exploring its applications, feel free to reach out. Collaboration and shared perspectives drive the field forward, enabling new breakthroughs in robotics.

The post Optimal Control of Compliant Robots: IEEE Transactions on Robotics Best Paper Award appeared first on David Braun.

Robotic intelligence is evolving beyond pre-programmed control, shifting towards systems that learn through real-world experience. A key challenge in this transformation is enabling robots to improve their performance on complex tasks—such as walking or riding a bicycle—without requiring highly accurate mathematical models or vast amounts of training data. This research introduces a novel concept: real-time optimal control for learner-helper robot pairs, where a learner robot refines its abilities through physical experimentation with the assistance of a helper robot. This approach bridges the gap between simulation and reality, unlocking new possibilities for autonomous learning in robotics.

Learner-Helper Pairs: A New Paradigm for Robot Training

At the core of this work is a robotic learning framework inspired by human skill acquisition. Just as a child learning to ride a bicycle benefits from an adult providing balance before riding independently, a learner robot receives structured assistance from a helper. The helper ensures the learner meets the minimum functional threshold—such as preventing a walking robot from falling—while gradually reducing support as learning progresses. This enables robots to refine their movements efficiently through repeated trials, without requiring highly accurate simulations or extensive datasets.

By shifting optimization from precomputed models to real-world experiments, this technique paves the way for new robotic control strategies that are adaptive, data-efficient, and capable of handling complex, unstable systems.

Key Applications of Learner-Helper Robotics

This research has broad applications across robotics and automation, including:

• Industrial Automation: Factory robots could learn tasks directly on the production line, improving efficiency without time-consuming reprogramming.
• Assistive Robotics: Medical robots and prosthetics could optimize their movements by interacting with users rather than relying on predefined control strategies.
• Autonomous Systems: Legged robots and exoskeletons could refine their locomotion in real-world environments, overcoming the limitations of rigid model-based control.

By removing the reliance on high-fidelity models and vast training datasets, this technique accelerates the adoption of intelligent, adaptable robots across industries.

Broader Impacts

Beyond specific applications, this research contributes to multiple domains:

• Bridging the Simulation-Reality Gap: Many control techniques struggle when transitioning from simulations to real-world implementation. The proposed approach ensures that robots optimize performance in real-time, adapting to uncertainties as they learn.
• Strengthening the Economy: Enabling robots to learn autonomously enhances manufacturing efficiency, reduces reliance on labor-intensive programming, and fosters technological innovation.
• Advancing Engineering Education: By integrating concepts from mechanical, electrical, and control engineering, this research promotes interdisciplinary learning and engagement.

Looking Ahead

The development of learner-helper robot pairs marks a step toward next-generation intelligent machines, where robots teach each other, refine their abilities through real-world trials, and adapt to uncertain environments. This work lays the foundation for more capable, autonomous robotic systems, pushing the boundaries of what machines can achieve.

This work was supported by the National Science Foundation Dynamics, Control, and System Diagnostics Program (Grant Number 2029181) under the Directorate for Engineering’s Division of Civil, Mechanical, and Manufacturing Innovation.

If this topic resonates with you, I would love to hear your thoughts or explore potential collaborations. Feel free to reach out and share your perspective.

The post Real-time Optimal Control of Engineering Systems: National Science Foundation Grant appeared first on David Braun.

Have you ever wondered how much faster humans could run with just a little help from physics?

A new twist on the old idea of spring-driven shoes may hold the answer. By storing and releasing energy during the swing phase of each stride, these wearables offer a fresh way to rethink running, potentially achieving speeds far beyond what is possible today. This concept is explored in my work, “How to Run 50% Faster Without External Energy“, published in Science Advances. The study explains how unpowered, spring-driven devices, such as robot boots, can enhance human locomotion by offering mechanical advantages similar to those of a bicycle.

The Skeptics’ Guide to the Universe hosted an engaging discussion about this work, delving into the science behind spring-driven wearables and the potential they hold for shaping the future of human-powered mobility. Listen to the panel’s insights in Episode #776:

The Physics Behind Spring-Driven Devices

The simplest model of these wearable devices is a variable-stiffness springs which stores energy while the leg is airborne and release it upon foot contact. This approach eliminates the “dead time” during a runner’s stride, enabling continuous energy use and significantly boosting efficiency.

Performance Potential

Physics suggests that speeds of up to forty-four miles per hour—comparable to top cycling speeds—are achievable under ideal conditions. Even practical designs could enable speeds of forty miles per hour, representing a fifty percent increase over the current limits of unassisted human running.

Applications

Potential use cases include:

• Sports: Enhancing athletic performance and creating opportunities for new competitive events.
• Emergency Response: Rapid mobility for law enforcement, search and rescue missions.

The reliance on mechanical systems rather than external power sources makes these devices especially suited for off-grid or resource-limited scenarios, while also supporting the goal of sustainable transport.

Challenges to Overcome

Key challenges include:

• Maintaining stability at high speeds.
• Minimizing risks of injury, such as strain or falls.
• Helping users adapt to the new movement mechanics required by the device.

Similar challenges were faced during the development of the bicycle. Early designs struggled with stability, safety, and usability before evolving into the efficient and widely used vehicles we know today.

Redefining Human Potential

Over the past century, technological advancements have revolutionized athletic performance. Innovations such as lightweight running shoes, aerodynamic swimsuits, and synthetic track surfaces have consistently pushed the boundaries of what athletes can achieve. Spring-driven devices may represent the next leap forward, offering unprecedented possibilities in speed and efficiency.

As we have seen in the evolution of the Olympic Games, technology does more than enhance performance—it challenges us to rethink the limits of human potential. Devices like these could one day become a standard part of athletic competition or recreational activities, paving the way for a future where humans achieve feats once thought impossible.

What are your thoughts on the future of spring-driven running devices? Could they redefine athletic performance and human mobility?

I would love to hear your perspectives—feel free to comment below or reach out if you are interested in discussing this further.

The post The Science Behind High-Tech Robot Boots appeared first on David Braun.

Can physics support running as fast as cycling? A recent news article explores spring-driven robotic exoskeletons, or Robo-boots, as a potential way to achieve this.

By storing and releasing energy efficiently, these boots mimic the mechanics of cycling to increase propulsion, potentially allowing humans to reach speeds of forty miles per hour. This approach highlights how physics principles can guide designs of wearable devices for enhanced mobility.

As the article explains, “Using a spring to store energy and release it is far more efficient than using your muscles to generate the same force.” This insight reflects ongoing efforts to combine biomechanics and robotics for enhanced human mobility.

A New Olympic Vision

Imagine a future Olympic Games featuring a new discipline: robot-enhanced sprinting. Athletes equipped with Robo-boots could test the limits of speed, precision, and endurance, creating a spectacle where human skill and robotic innovation intersect. The challenge would no longer be solely physical fitness but also how seamlessly athletes adapt to and control cutting-edge technology, pushing the boundaries of what humans can achieve.

Such a sport could inspire advancements in robotic exoskeletons and set new standards for safe, efficient mobility systems, impacting not just athletics but broader applications in daily life and work environments.

Challenges in Development

The challenges faced by Robo-boots are similar to those encountered during the early days of the bicycle. Ensuring safety, comfort, and widespread usability requires further development. Lightweight designs and intuitive integration with human movement remain priorities for research.

An Invitation to Explore

It is exciting to consider how innovations like Robo-boots might integrate into everyday life. Whether enhancing performance, saving time, or reducing effort, this concept invites thoughtful discussion about the relationship between technology and human potential.

Explore the full article on The Conversation for a deeper dive into this concept.

If this topic resonates with you, feel free to reach out to discuss the potential and ideas behind this technology.

The post Robo-boot Concept Promises 50% Faster Running appeared first on David Braun.

Human mobility has long been shaped by biomechanics and physics, but new research suggests that running speeds could far exceed current records with the right mechanical assistance. Inspired by physics-driven models, we are exploring a new concept of a spring-driven running device that could, at least in theory, allow runners to reach speeds once thought impossible—potentially up to 50 miles per hour, far beyond the limits of unassisted human sprinting.

The Science Behind Spring-Powered Running

At the core of this innovation is a fundamental shift in how energy is used during running. Typically, human runners generate power primarily when their feet push off the ground. However, computer models suggest that much greater energy efficiency is possible if runners can also exert force while their feet are in the air.

Given this insight, we are exploring the feasibility of a spring-driven wearable device that stores and releases energy during the swing phase of a runner’s stride. By optimizing this phase, the device could eliminate wasted time in each step and dramatically improve speed and efficiency.

Potential Applications of Spring-Driven Running Devices

The proposed concept has broad implications, including:

• Athletics and Competitive Sports: Running-assisted technology could redefine the limits of human performance, introducing new categories in competitive running and even reshaping how elite athletes train.
• Emergency Response and Search & Rescue: Faster, more efficient movement could provide firefighters, law enforcement, and rescue personnel with new tools for rapid response in critical situations.
• Military and Tactical Use: Mobility-enhancing technology could give soldiers greater endurance and speed in combat zones, reducing physical strain and improving operational effectiveness.

By leveraging the power of mechanically adaptive wearables, these devices open new possibilities for human-powered transportation, offering an energy-efficient alternative to vehicles in unstructured environments.

Challenges and Future Development

While the concept is promising, several challenges must be addressed before spring-driven running devices become mainstream:

• Stability at High Speeds: Maintaining balance and coordination while running at unprecedented speeds requires careful design and adaptive control systems.
• Injury Prevention: Increasing energy input could introduce new biomechanical risks, such as excessive joint strain or falls.
• Adaptation to New Motion Mechanics: Just as early bicycles required practice to master, runners will need time to adjust to new movement patterns.

Redefining Human Performance

Throughout history, technology has consistently pushed the boundaries of human capability—from lightweight running shoes to aerodynamic cycling gear. Spring-driven wearables could mark the next major leap, allowing humans to run faster than ever before while using physics to their advantage.

Explore the full article on The Guardian.

If this vision of enhanced human mobility excites you, I would love to hear your thoughts. Could spring-driven wearables redefine the future of running? Feel free to comment or reach out to discuss!

The post Spring-driven: Concept that could see Usain Bolt rocket to 50mph appeared first on David Braun.

Welcome to the 19th IEEE/RAS-EMBS International Conference on Rehabilitation Robotics (ICORR 2025), the biannual conference on theoretical and experimental issues in the fields of Rehabilitation Robotics and Neuroscience. ICORR covers several disciplines with both theoretical and experimental challenges in robotics, control engineering, and neuroscience applied to healthcare. Focuses include cutting edge solutions to boost the rehabilitation process, providing robotic assistance to address and speed motor recovery, and trying to unveil mechanisms underlying brain plasticity. Enabling better societal inclusion, as well as improving independence and quality of life, are the main goals we aim to achieve, with the purpose of providing health systems with technological instruments to assist human beings in prevention, diagnosis, rehabilitation, and personal assistance.

The post Associate Editor at the IEEE International Conference on Rehabilitation Robotics appeared first on David Braun.

The 10th iteration of the IEEE International Conference on Biomedical Robotics and Biomechatronics – BioRob2024 – represents a collaborative endeavor between two IEEE Societies: Robotics and Automation (RAS) and Engineering in Medicine and Biology (EMBS).

BioRob delves into both theoretical and practical challenges arising from the integration of robotics and mechatronics into medicine and biology. The main objective of Biorobotics is to evaluate biological systems through a ‘biomechatronic’ lens, striving to unravel the scientific and engineering principles that underpin their exceptional performance. This deep comprehension of biological system functions, behaviors, and interactions serves two primary purposes: to inform the design and creation of new, high-performance bio-inspired machinery and systems for various applications, and to foster the development of innovative nano-, micro-, and macro-devices that can act upon, replace, or assist human beings in areas such as disease prevention, diagnostics, surgical procedures, prosthetics, rehabilitation, and personal assistance.

The post Editor at the IEEE BioRob Conference appeared first on David Braun.

The post Exoskelette: Schneller dank Sprungfedern appeared first on David Braun.

The post Area Chair for the 2018 Robotics Science and Systems Conference appeared first on David Braun.

The post Singapore Ministry of Education Academic Research Fund appeared first on David Braun.

Speed ​​records have a strange, primal allure. If an angry elephant is chasing you, would you be fast enough to escape? What about a bear or a black mamba? In any of these cases, your chances are not very good unless you are a trained runner. Unless you had springs on your legs. What sounds like an idea from an animated series could actually work. That’s the conclusion of a new study conducted at Vanderbilt University. Read the full article here.

The post Running Shoe with Programmable Spring appeared first on David Braun.

One way to imagine human augmentation is to consider an energetically active robot that supplements muscle power with external energy. Another way to imagine human augmentation is to consider solely human-powered energetically passive devices; two examples are bicycles and ice skates. In this talk, I will discuss the hidden potential of an emerging human-powered compliant actuation concept, and its potential to enable the creation of energetically passive human augmentation devices. Read more about the event and the speakers here.

The post Keynote at 2020 International Conference on Intelligent Robots and Systems Workshop appeared first on David Braun.

While state-of-the-art compliantly actuated robots are able to achieve previously unprecedented agility, robustness, and adaptability, they operate with low efficiency. Energy-efficient compliant actuators are missing ingredients, and key enablers of next-generation robotic  systems: prosthetic devices, orthotic devices, and wearable exoskeletons. However, despite recent advances in controlling these variable stiffness systems, their design remains experience-based and not well understood. This presentation will cover our recent effort in developing analytical and optimization-based frameworks for the design of intrinsically efficient variable stiffness systems.

The post Invited Speaker at the ICRA 2017 Workshop: Mechanics of Human Locomotion and the Development of Wearable Robotic Systems appeared first on David Braun.

The benefit of using dynamic optimization for the control of industrial robots has been recognized by the robotics community. However, many of the methods for dynamic optimization use heuristic treatment of physical constraints, limitations on the control inputs, and constraints on the temporal aspect of the motion, common in applications. This talk presents our recent effort to develop a numerically efficient dynamic optimization method which takes physical constraints rigorously into account. The method complements a typical model-based planning approach with an online optimal constrained feedback controller. The method is suitable for real-world application under imperfect model information. Read more about the event and the speakers here.

The post Invited Speaker at the ICRA 2017 Workshop: Recent Advances in Dynamics for Industrial Applications appeared first on David Braun.

The post Scientific Program Co-Chair of the International Conference on Rehabilitation Robotics appeared first on David Braun.