Robots of the
future human (2.0)

Our exoskeleton portfolio consists of near-production platforms currently in advanced development, with core subsystems validated through initial functional and integration testing. These systems are engineered as modular human-augmentation solutions optimized for reliability, ergonomics, and scalable deployment. From a system configuration perspective, the platform family is structured in two main classes. Full-body exoskeletons provide coordinated actuation and biomechanical alignment across the entire musculoskeletal chain, enabling comprehensive load redistribution, endurance enhancement, and motion assistance. Partial exoskeletons are designed as targeted modules focusing on specific anatomical regions—such as lumbar support units, upper-limb assist modules, or lower-extremity actuation frames—allowing application-specific optimization with reduced system complexity. In terms of use cases, the platforms are being developed across three primary operational domains. Medical configurations focus on rehabilitation, mobility restoration, and assistive therapy workflows with emphasis on safety, compliance control, and patient adaptability. Industrial and daily assistance variants are optimized for fatigue reduction, injury prevention, and sustained performance in physically demanding environments. Defense-oriented configurations explore high-endurance and load-augmentation capabilities for mission scenarios where operator efficiency and resilience are critical. Across all variants, the design approach emphasizes modular hardware architecture, sensor-driven control loops, and extensible software frameworks, enabling iterative refinement as the systems progress toward full validation and deployment readiness.

Core Technology

At the core of our platforms lies an integrated technology stack that combines advanced artificial intelligence with next-generation materials engineering. Our control architecture leverages adaptive AI models capable of learning user-specific motion patterns and biomechanical responses in real time, enabling predictive assistance rather than purely reactive actuation. Sensor fusion pipelines—integrating EMG, IMU, force, and kinematic data—feed into intelligent control loops that dynamically optimize torque distribution, stability, and energy efficiency across different movement phases. In parallel, we develop and implement high-performance structural and functional materials to enhance durability, weight efficiency, and ergonomic compliance. This includes lightweight composite load paths, energy-return elastic elements, and smart interfaces designed to improve human–machine coupling. By co-designing hardware, materials, and AI-driven control systems, we are building exoskeleton platforms that are not only mechanically capable but also context-aware, adaptable, and ready for real-world deployment.
Our platform integrates a multi-axis inertial measurement unit array with surface electromyography sensors, enabling intent detection before movement onset. Actuation is achieved through a hybrid hydraulic-electric drive system, delivering high torque at low noise levels.

Specifications (Development Mode)

Achievements and Reached Level

The first domain focuses on human power amplification, aiming to augment the physical capabilities of healthy individuals. This includes enabling users to perform demanding tasks—such as carrying heavy industrial loads, assisting in disaster response operations, or enhancing endurance in field conditions—with reduced fatigue and increased efficiency. In addition, the system is designed to provide adaptive robotic support for everyday activities when needed. The second domain addresses rehabilitation and assistive technologies for individuals with partial or complete disabilities, whether temporary or permanent. In this context, our systems support recovery processes, assist mobility, and provide intelligent guidance during rehabilitation—for example, enabling gait training after neurological injuries or supporting upper-limb motion in post-surgical recovery. Across both domains, we are developing an intelligent control architecture capable of operating both partial (limb-specific) and full-body exoskeleton systems. Our approach emphasizes tight integration with the human neuro-musculoskeletal system, allowing for more adaptive, robust, and resilient behavior compared to conventional exoskeleton designs. The results we have achieved so far demonstrate strong progress. Building on this foundation, we continue to advance toward a highly responsive and controllable system that can seamlessly interact with human physiology. Furthermore, we are actively enhancing the core robotic platform to support higher-level applications—including military operations, elite athletic performance, and search-and-rescue missions—where precision, reliability, and real-time adaptability are critical.

Vision and Impact

Our long-term objective is to transform exoskeletons from specialized equipment into indispensable intelligent companions that meaningfully expand human capability and independence. We envision assistive configurations becoming everyday mobility tools for individuals with disabilities—enabling activities such as autonomous commuting, household interaction, and social participation with minimal external support. In parallel, strength-augmentation variants are designed to provide measurable load-handling advantages for soldiers and heavy-duty workers, improving endurance, reducing injury risk, and extending operational effectiveness in physically demanding environments. Another key milestone is the development of remotely configurable and supervisable systems tailored for aging populations, allowing caregivers or clinicians to monitor performance, adjust assistance profiles, and ensure safety in real time. Complementing the hardware layer, our Agentic AI–driven lifelong assistance framework aims to evolve into a persistent cognitive layer that learns user routines, anticipates needs, and continuously optimizes support strategies. Beyond product outcomes, we also see this program as a catalyst for generating deep R&D capabilities—advancing human augmentation science, intelligent control methodologies, and translational robotics knowledge that can power future innovations across multiple industries.

Next Steps and Boosting

We also work on artificial muscles to replace the motor technologies with more robust mechanisms inspired from biological muscular system

From human body to a mice, a bird a dinosaur if still exists... Specialized cascaded neural network for body interpretation. Using neuromorphics. The SABICRAFT-Arbon Humanoid research and development is a full-body autonomous robotic platform engineered for dynamic environments requiring dexterous manipulation, stable bipedal locomotion, and natural language-driven task execution. Designed from first principles with embodied intelligence in mind, it represents our most ambitious system to date.

Locomotion & Balance

Balance control is achieved through a whole-body model predictive control (MPC) framework running at 1 kHz, drawing from 42 degrees of freedom across joints with torque sensing in all actuated links. The system maintains stability on uneven terrain, slopes, and in response to unexpected perturbations.

Intelligence Layer

Task planning and natural language understanding are handled by a fine-tuned multimodal model running onboard, enabling the system to receive spoken or typed instructions and decompose them into primitive motion sequences. Long-horizon planning with memory recall is an active research focus.

We design and build next-generation embodied robotic systems—humanoid, animatronic, and beyond—where advanced artificial intelligence meets precision engineering. As breakthroughs in AI, high-performance electronics, and material science converge, the boundary between digital intelligence and physical capability is rapidly dissolving. This transformation is not incremental; it is foundational. Robust, adaptive, and intelligent machines are no longer theoretical—they are achievable, scalable, and ready to redefine how humans interact with technology in real-world environments. Our expertise lies in translating this technological momentum into reliable, high-performance systems. By integrating state-of-the-art control architectures, intelligent perception, and durable mechanical design, we develop embodied platforms capable of operating with precision, resilience, and autonomy. We approach each system as a synthesis of intelligence and form—engineered not only to function, but to perform at the highest standards of modern robotics.

Embodiment

Our focus embodied robotics is intelligent systems where cognition is inseparably integrated with a physical form—machines that do not merely compute, but perceive, act, and adapt within the real world. Unlike disembodied software, embodied intelligence emerges through continuous interaction between algorithms, sensors, actuators, and the environment. This paradigm marks a decisive shift: intelligence is no longer confined to data centers, but realized through motion, force, and presence—enabling machines to learn from and respond to the complexity of physical reality. In the new generation of robotics, embodiment is the key to unlocking true autonomy and utility. Advances in AI, edge computing, and advanced materials have made it possible to engineer systems that are not only intelligent, but context-aware, resilient, and physically capable. Embodied robotics represents the convergence of mind and machine—where digital reasoning meets real-world execution—paving the way for robots that can seamlessly collaborate, assist, and operate across industries with unprecedented sophistication.

Soft Materials and New Generation Actuators

We have been working on artificial muscles and tendon-like attachers since 2016

Since 2016 we have been working on research and development of next generation exoskeleton and hybrid models for human power amplification, daily assistance support units for disabled people and medical rehabilitation

There are 2 main projects we continue :

SABICRAFT-ARBON as Exoskeleton
Intelligent Controller Unit for existing Robotic Units
Artificial Muscles and Soft Robotics
Generalized Neural Network model for Body-type mechanisms

Purpose and Industries

Body Assistance for Disabled or Elderly People
Rehabilitation Unit for Medical Purpose
Emergency - Rescue
Manufacturing - Factories - Warehouses
Autonomous Vehicles as Robotic Units