Neura Robotics CEO David Reger on Cognitive Platforms, Humanoid Design, and the Path to Physical AI

Neura Robotics CEO David Reger Offers Insights into Cognitive Platforms and Humanoid Futures

David Reger, founder and CEO of Neura Robotics, recently joined Marwa ElDiwiny and Scott Walter for an in-depth podcast interview, offering a detailed look into the company's philosophy, its journey from a 2019 startup to a significant player in cognitive robotics, and the intricate engineering behind its expanding robot lineup, including the much-anticipated 4NE-1 (pronounced “for anyone”) humanoid. Neura is set to reveal a new version of 4NE-1 at the Automatica show in Munich, June 24-27, an event co-host Scott Walter mentioned he will be attending, with the possibility of a follow-up live interview from the show.

The conversation comes on the heels of Neura's significant announcement of a collaboration with SAP and NVIDIA to integrate business AI with physical robots, a topic that underscores Neura's broader ambitions. Reger's discussion provides valuable context to these strategic moves, emphasizing a long-term vision rooted in creating truly cognitive machines.

Beyond Hardware: A Platform for Cognitive Robotics

A core theme emerging from the interview is Neura Robotics' foundational commitment to building more than just advanced robotic hardware. Reger elaborated on his initial vision in 2019: to imbue robots with cognitive capabilities – the ability to see, hear, feel, and make decisions – effectively creating "physical AI." This wasn't about adding another robot to a crowded market, but about developing a comprehensive platform, the Neuraverse, designed to enable robots to learn, adapt, and be easily integrated into various environments.

"The idea was...making robots cognitive," Reger explained, highlighting the goal to move beyond complex programming to intuitive interaction. He describes the Neuraverse as an ecosystem where skills and applications can be developed and shared, accelerating the deployment and utility of robots across industries. This platform-centric approach is central to Neura's strategy, aiming for a sustainable model where the community can contribute to the evolution of robotic capabilities, rather than constantly "reinventing the wheel."

Neura Robotics' intelligent palletizing solution in action. This application highlights how their cognitive robotics platform, the Neuraverse, supports a diverse range of hardware – from cobot arms to humanoids – to bring adaptable automation to various industrial tasks.

Deep Engineering: The Nuts and Bolts of 4NE-1 and Beyond

The interview also delved into the significant in-house engineering efforts at Neura, particularly concerning their humanoid robot, 4NE-1. Reger discussed the iterative design process for critical components like actuators. He revealed that Neura develops its own motors, encoders, brakes, and even innovative gearbox solutions tailored for the demanding requirements of humanoids, such as robustness against falls and high power density in compact forms.

For instance, Reger mentioned advancements in their actuators, achieving more than double the torque (from around 150 Nm to over 300 Nm) in the same size, and the development of new cycloidal-based gearboxes and inductive encoders to meet the unique challenges posed by humanoid locomotion and interaction. This vertical integration, while complex, gives Neura precise control over the performance and reliability of its systems.

The design of the robot's hands was another area of focus, with podcast co-host Scott Walter describing them as the "business end" of the humanoid. Neura is progressing from an earlier hand design, which Reger referred to as having six degrees of freedom (DoF), to a more dexterous 16-DoF version, with even more advanced designs in the research phase. During the podcast, Walter clarified that in the context of the earlier hand, 6-DoF referred to six actuators controlling the hand's movements, though the hand itself possessed more than six physical joints. This distinction highlights the engineering complexity in achieving fine motor control. The progression towards a 16-DoF (actuator) hand underscores the critical role of manipulation in achieving versatile, human-like capabilities. The pursuit of enhanced robotic dexterity is a critical area of research across the field. Yesterday, we covered another significant initiative where a UK research team secured £5 million to advance human-like robot dexterity through innovative AI co-design methods.

The Humanoid Form Factor and Safety

Addressing the perennial "why humanoid?" question, Reger pointed to the human-centric design of our world and the intuitive understanding the humanoid form offers to people, potentially accelerating adoption. However, he also emphasized a pragmatic approach, noting that Neura supports various form factors, including wheeled robots like the recently announced MiPA, until humanoid technology matures sufficiently in terms of stability, cost, and overall effectiveness for all applications.

Beyond the factory floor: Neura Robotics' MiPA showcases their exploration of personal robotic assistants. Such applications align with CEO David Reger's vision of a versatile cognitive platform enabling robots to learn and assist in diverse human environments.

Safety remains a paramount concern. Reger, himself a safety engineer, detailed Neura's focus on redundant and diverse safety systems. He highlighted their development of a "Safe Human Detection (SHD) sensor," aimed at enabling robots to identify and avoid contact with humans proactively, a crucial step towards true collaboration.

Hear It From The Source

While this article touches on some key themes, the full interview with David Reger offers a much richer and more nuanced understanding of Neura Robotics' vision and technological endeavors. Listeners can gain deeper insights into the company's strategic thinking on AI integration, the challenges of building robust and scalable robotic solutions, and Reger's perspective on the future of human-robot interaction.

The full discussion is highly recommended for anyone following the rapidly evolving field of robotics and intelligent automation. Listen on Spotify or get the full conversation on YouTube here: