A primer on investing in neurotechnology

Neurotechnologies are a category of solutions that are widely discussed yet still significantly underappreciated. These technologies are designed to monitor, interpret, and even communicate with the nervous system, while reducing friction in human-machine interaction by enabling direct access to the body’s most natural and efficient information transmission network. In my view, advances in neurotechnology may become one of the most important technological developments of our time, perhaps comparable in impact to Artificial Intelligence (AI).

But what exactly are neurotechnologies? Are they science fiction? Can mind control, like the technology portrayed in the Black Mirror episode “Plaything,” directly interpret emotions, memories, and perceptions from the brain? In reality, neurotechnologies encompass a broad range of technologies developed to connect with, interpret, stimulate, or interact with the human nervous system. Among these are human-machine interfaces and Brain-Computer Interfaces (BCIs), which can translate neural signals into computational language and enable communication with and interpretation of brain activity. These technologies are opening new possibilities for diagnosis, rehabilitation, and everyday assistance. Another important field is neuromodulation, which uses electrical, magnetic, or ultrasound-based stimulation to modulate nervous system activity, supporting pain management, motor rehabilitation, and the treatment of various neurological conditions.

When viewed through this lens, success in neurotechnology has the potential to profoundly transform the human condition. These technologies could help treat movement disorders, paralysis, and other neurological conditions, while also expanding capabilities related to memory, creativity, and complex thinking. Beyond healthcare, they may drive an unprecedented transformation in both the economy and the future of computing. In this article, I present an overview of the neurotechnology landscape, with a particular focus on Brain-Computer Interfaces (BCIs) and neuromodulation systems, as well as the key market trends and criteria for evaluating the growth potential and long-term evolution of this emerging sector.

This is what enables movement. But movement is not just the result of a signal being sent.

Your nervous system operates as a continuous control system, that is, a system that continuously monitors outcomes and adjusts actions based on feedback. It monitors what is happening, compares it with what should happen, and adjusts in real time. It refines force, corrects errors, and reacts instantly to unexpected changes, like a slipping object. All of this happens in milliseconds, without conscious effort. In other words, your body is not simply executing commands. It is constantly sensing, processing, and adapting.

This is physical common sense, a form of closed-loop intelligence in which actions are continuously adjusted based on feedback. Built through experience, it enables the body to handle uncertainty, refine actions, and recover from disturbances in real time.

And the point is, it is electrical – and electrical signals can be measured, modulated, and applied.

Modulating Neurons, Generating Movement

Because of this, it becomes possible not only to understand how the system works, but to interact with it. One approach is Functional Electrical Stimulation, which uses small electrical pulses to generate movement in the body (Khan et al., 2023). It has been widely used in people with paralysis or muscle weakness to restore functions such as grasping, walking, and posture. In practice, electrical stimulation need not recreate the entire system. It simply activates it. Applying an electric current through electrodes stimulates motor neurons that control the muscles. When the stimulus reaches a certain level, it generates an action potential in these neurons. This signal travels along the nerve to the muscle, triggering contraction; as a result, movement occurs (Crago et al., 2021; Kandel et al., 2014).

But electrical stimulation does more than just generate movement. It also activates sensory pathways, sending signals back to the brain. This feedback helps the nervous system adapt, learn, and reorganize, a process known as neuroplasticity ( Crago et al., 2021; Sousa et al., 2022). Because of this, electrical stimulation is not limited to rehabilitation. Even in healthy individuals, it can be used to better understand movement, refine control, and improve performance. This means electrical stimulation is not just about activating muscles. It is about directly interacting with the neural signals that generate movement.

Why Electrical Stimulation Alone Struggles to Produce Functional Movements

The human body does not generate movement through isolated signals. Your nervous system operates as a continuous control system. It monitors what is happening, compares it with what should happen, and adjusts in real time. It refines force, corrects errors, and reacts instantly to unexpected changes.

Movement, therefore, is not a single event. It is a process.

Take a simple action like picking up a glass. It involves multiple coordinated phases: reaching, adjusting the hand, applying the right amount of force, stabilizing the object, and correcting for any small slips. Each phase depends on continuous feedback and adaptation. The same applies to more complex actions, such as walking. Gait is not just muscle activation. It is a sequence of phases coordinated over time, like an orchestra. Each muscle group must activate at the right moment, with the right intensity, and in coordination with others. This is where electrical stimulation, on its own, faces important limitations.

Traditional electrical stimulation systems often operate in an open-loop manner. In simple terms, an open-loop system sends a signal but does not check what actually happened. It does not adjust based on the result. It is like pressing a button and assuming everything worked as expected.
In contrast, the human nervous system works as a closed loop. It constantly senses what is happening, compares it with what should happen, and adjusts continuously. It is more like driving a car, where you constantly correct the steering based on the road.

While open-loop stimulation is simple and enables fast responses, it lacks the ability to adjust stimulation in real time to the user’s needs (Dong et al., 2022). As a result, stimulation can generate muscle contraction, but may struggle to achieve the coordination, timing, and adaptability required for fully functional movement. In other words, electrical stimulation can produce movement, but not always movement that is precise, adaptive, and naturally coordinated.

The missing element is the same one that defines natural movement: a closed-loop system that continuously senses, adapts, and coordinates actions over time. Without this, stimulation remains primarily a trigger rather than a fully adaptive controller.

Phases of human gait are coordinated in a continuous cycle of adaptation, stability, and motor transition.

Data can create controlled movement if it is the right data

The challenge of reproducing adaptive movement lies in building systems that can infer structure from complex data and generalize beyond predefined behaviors. But language models have shown us something important: when trained on large, diverse data, systems learn patterns that generalize across situations. The same is true for movement. The nervous system does not control the body with fixed commands, but through a continuous feedback loop. The brain defines a goal, and the body adjusts in real time using movement-related signals, such as position, muscle activity, and force (Dong et al., 2022). Movement, therefore, is not executed as a predefined sequence; it is continuously controlled.

This is the central challenge in Functional Electrical Stimulation. Most systems attempt to reproduce movement through predefined trajectories, but the human body does not follow fixed scripts. It changes constantly as muscles fatigue, conditions shift, and responses vary. As a result, fixed control strategies tend to fail in the presence of variability, necessitating adaptive approaches.

At Orby.co, we use artificial intelligence to move from imitation to control. Our portable device captures biomechanical signals, measurements that describe movement in real time, and integrates them with reinforcement-learning-based control, a method in which the system learns to adjust its actions based on feedback from the environment. Instead of replaying movements, the system continuously adapts how it controls them. It observes the body, adjusts electrical stimulation, and updates control at every moment, with sensors maintaining the loop closed and enabling real-time responses to changes in the body.

This is how electrical stimulation becomes functional movement. To enable this in practice, we use small models that run directly on the device, allowing fast, low-latency, real-time control. The result is a clear shift in behavior: movements become smoother, more natural, and more stable, as the system reacts continuously rather than following predefined commands. Our goal is not to copy movement, but to continuously control it.

Developed by Orby.co, the Ortech platform integrates biomechanical sensors, AI, and real-time adaptive electrical stimulation.

Early signs of physical intelligence

Early signs of physical intelligence emerge as real-time interaction with the human body is scaled, revealing that systems capable of continuous modulation adapt faster, handle variability more effectively, and produce more fluid and functional movement. This capability, however, depends on continuous sensing. At Orby.co, the approach is grounded in capturing biomechanical signals that reflect movement state in real time and using them to adjust electrical stimulation in real-time. This shifts control from targeting isolated joints or predefined trajectories to regulating movement as an integrated, dynamic process that is continuously updated based on the body’s current state.

In this framework, generalization emerges from interaction. The same system can support multiple movement tasks, such as walking, grasping, and stabilization, without requiring task-specific redesign. It is not restricted to a single body region, such as the arms or legs, but operates across the body because it is based on underlying movement dynamics. At the same time, the body itself is not constant. Skin impedance, or the effective resistance of the skin to electrical current, and neuromuscular responses vary with fatigue (a reduction in the muscle’s ability to generate force), posture, and physiological condition, introducing variability that fixed controllers cannot accommodate.

By modulating stimulation in real time, the system adapts to these changes, enabling adjustment across different body dynamics, individuals, and skin impedance conditions. This allows both generalization across tasks and personalization to individual characteristics. As a result, control becomes fluid, with smoother movements, natural transitions, and continuous adjustment in response to variability. Rather than operating as a fixed controller, the system functions as a real-time adaptive control framework.

Movement is more than low-level control

Electrical stimulation has traditionally been used to activate muscles through predefined parameters such as timing and amplitude. However, the human body behaves as a nonlinear, time-varying system that cannot be reliably controlled through fixed commands. That is, its response is not proportional to inputs and changes over time depending on internal and external conditions.

When stimulation is combined with real-time sensing and adaptive control, movement becomes a continuous interaction rather than a predefined sequence. Biomechanical signals, as real-time representations of movement, enable dynamic modulation of stimulation, compensating for variability such as fatigue, posture, and changes in impedance.

Evidence from multiple studies supports this shift. Reported outcomes include improvements in biomechanical movement parameters, particularly joint kinematics, such as increased range of motion in the hip, knee, and ankle. These effects are especially evident in hip and knee flexion and ankle dorsiflexion when compared to conventional reflex-driven Functional Electrical Stimulation strategies. In addition, these approaches have been associated with enhanced motor relearning and the promotion of neural plasticity.

At Orby.co, this paradigm enables control across multiple body regions and movement types without task-specific programming. Continuous modulation based on real-time sensing allows both generalization across conditions and personalization to individual dynamics. For example, in an upper-limb control protocol, the system modulates movement parameters such as joint angle, velocity, and trajectory to enable controlled elbow flexion and extension. Instead of relying on fixed stimulation patterns, stimulation is continuously adjusted in real time based on feedback, enabling smooth, coordinated transitions between movement phases.

This type of control is essential for functional actions such as reaching, bringing objects toward the body, and positioning the hand in space, which depend on continuous adjustment of movement rather than discrete activation. In this context, functional movement emerges from real-time adaptive control, shifting Functional Electrical Stimulation from muscle activation to continuous motor regulation.

Application of functional electrical stimulation for neuromuscular modulation in the arm.

References

Hamrah, A. (2024). A primer on investing in brain-computer interfaces. Substack. https://hamrah.substack.com/p/a-primer-on-investing-in-brain-computer.
Grand View Research. Brain Computer Interface Market Size | Industry Report, 2033 [Internet]. San Francisco: Grand View Research; 2025 [cited 2026 Jun 3]. Available from: https://www.grandviewresearch.com/industry-analysis/brain-computer-interfaces-market/
Rao, N. (2025, March 10). 2024 neurotech funding snapshot. Neurotech Futures. https://neurotechnology.substack.com/p/2024-funding-snapshot