Cover Letter for the Humanoid BCI-controlled Robot Proposal

At the moment, there is an avalanche of growth in technologies used to create humanoid robots. In part it is due to the rapid development in microelectronics, sensor systems, new materials and software algorithms, etc. It is also related to public expectations that have been reinforced by the long-term predictions found in science fiction.

There are pressing practical problems that require the creation of an anthropomorphic robot: Rehabilitation of patients with skeletal-muscular and nervous system disorders, lost limb and organ replacement, and exoskeletons to provide mobility for people with disabilities.

A second group of practical problems requiring creation of an anthropomorphic robot is related to human safety when operating in a hostile environment. Work in areas contaminated with chemicals or radiation, work in airless environments, disposal of explosive devices, and military operations will become far more effective and safe when conducted via remote-controlled humanoid robot. Creation of such a robot is especially significant for its use in near-Earth space, or for the study of the planets in our Solar System.

Further, there are a significant number of applications for a humanoid robot which are not so vitally important, but will be supported by consumer demand: Domestic use, sports, and in the entertainment industry. Furthermore, the results of research and development of robot's design elements, new materials and software algorithms, and integration of individual components will be useful to industry before the robot design is complete and ready for production.

Humanoid robot development has been in progress for at least two decades. American and Japanese engineers are the most successful in this field. Russian developments are slightly lower than the global level, but nevertheless, there are certain works that remain relevant and in some ways outperform the global standard.

Therefore, it is important to create a BCI-controlled humanoid robot in Russia because of:

1. The necessity to evolve and implement existing developments.

2. The need to issue a directive for local scientists to research and develop the missing elements of the robot.

3. The opportunity to realize the nation’s potential in science and engineering through enhanced international cooperation.

In addition, the development of production capabilities for manufacturing the robot's components, as well as its full structural elements, will create a new industry providing a significant number of jobs for skilled workers and entrepreneurs in Russia.

We propose the creation of a humanoid robot controlled via "brain-computer" interface, reading a human operator's EEG parameters. It is important and necessary to create a support system for the robot's autonomous operation: an artificial nervous system allowing the robot autonomous behavior outside of human operator's direct control (maintaining balance, motor function of limbs, etc.), with the intent of eventually increasing the robot’s capacity for autonomous action to the point that the machine will act on its own to complete goals set by the operator.

Creating a control system for a humanoid robot is not a trivial task. The range of possible uses for such a robot is very extensive and it demands a wide range of remote control capabilities: From providing nearly complete autonomy that requires only that the operator set goals, to controlling the robot's smallest actions or even movements in real time.

Conventional interfaces used for remote control of technical devices and systems are unable to provide the operator with the possibility of simultaneous control of all degrees of freedom in a humanoid robot, in real time. Indeed, the number of degrees of freedom may be close to that of a human body, even considering the fact that basic movement organization (maintaining balance, posture, individual motor synergy) can be automated.

In addition, some applications of the robot imply that it is controlled by a completely paralyzed operator who is able to neither generate voice commands, nor use traditional interfaces like the keyboard or control panel. Although neural interfaces currently provide less throughput than the traditional ones, we believe that they have the potential to become exactly the means to control the robot and achieve our goals.

At the moment, some specialists and research groups collaborating with the "2045" Strategic Social Initiative have developed devices and components that are potentially useful for integration into a BCI-controlled robot construction. We believe that the general problem statement forms an integrative framework for a number of R&D areas, opens new perspectives, and starts new scientific trends. For example, problems in the areas of psycholinguistics of altered states of consciousness and of biofeedback acquire new meaning in the context of creating a universal BCI able to provide biocontrol of robot whose number of degrees of freedom is closer to that of a human body.

The project consists of three main areas of research and development. Each includes the following projects:

1. Create structural elements of the body.

2. Develop autonomous behaviors for the body.

3. Develop a system of external control via neural interface.

I. Creating the structural elements of the body involves:

1. Creation of a body and its power components -- an arm, a hand, a bipod that includes the robot's legs, and a model head with facial features.

• Vladimir A. Konyshev, President of Neurobotics.

Development of animatronic head with the elements of communicative and emotional behavior.

• Mikhail Y. Yablokov, Ph.D. in Physical and Mathematical Sciences, Senior Researcher at the Heat-Resistant Thermoplastics Laboratory at ISPM (Russian Academy of Sciences).

Creation of electro-active polymeric structures for the "artificial muscles" using thin film technologies.

• Professor Alexander A. Frolov, PhD. in Biology, Chief of the Mathematical Neurobiology Learning Lab at the Institute of Higher Nervous Activity and Neurophysiology (Russian Academy of Sciences).

Creation of a humanoid robot and a neural network controller that ensures the sustainable management of robot's movements through both direct communication and feedback.

• Vladimir A. Konyshev, Head of Neurobotics, Robotics Research and Training Center at the Bauman Moscow State Technical University.

Creation of an anthropomorphic leg.

• Mikhail Y. Eingorin, Head of the Computer Problems Laboratory, Head of the Department (1964 -1990) and Assistant Professor at Lobachevsky State University of Nizhni Novgorod, President of SKIT R&D and Pilot Design Company.

Development of digital linear and circular engines to provide the motor capacities of the robot.

2. Creation of a sensory system (sensory organs): Sight, hearing, organs of chemoreception (smell and taste), artificial skin (capable of discerning texture and temperature). In the long term we expect to empower the robot with sensitivity to IR, UV, and magnetic fields, extending the sensory capabilities of a human operator.

• Professor Vladimir G. Yakhno, Ph.D. in Physical and Mathematical Sciences, Head of the Auto-wave Processes Group, Laboratory Chief at the Institute of Applied Physics (Russian Academy of Sciences).

Creation of a system to rapidly analyse liquid multi-component products, corresponding to a given standard.

• Mikhail Y. Yablokov, Ph.D. in Physical and Mathematical Sciences, Senior Researcher of the Heat-Resistant Thermoplastics Laboratory at ISPM (Russian Academy of Sciences).

Creation of a multi-sensory analyzer of olfactory data - an “electronic nose".

3. Development of the robot’s power supply.

II. Development of autonomous behavior includes the following projects:

1. System of internal coordination: The system of maintaining equilibrium at rest and in motion, and a system to track objects in the external world.

• Professor Alexander A. Frolov, PhD. in Biology, Chief of the Mathematical Neurobiology Learning Lab at the Institute of Higher Nervous Activity and Neurophysiology (Russian Academy of Sciences).

Creation of a humanoid robot and a neural network controller that ensures sustainable management of the robot's movements through both direct communication and feedback.

• Valery D. Zuckerman, Applied Research deputy director at the Neurocybernetics Research Institute of South Federal University, Head of the Laboratory of self-organizing brain networks.

Creation of adaptive self-adjusting body control systems.

• Professor Vladimir G. Yakhno, Ph.D., Head of the Auto-wave Processes Group, Laboratory Chief at the Institute of Applied Physics (Russian Academy of Sciences).

Creation of a system to analyse images from the set of sensors and select the objects of the specified class, matching their description in the operator’s language. Oriented for “parsing debris”.

2. Systems to harmonizing with the external environment: The system to mimic patterns and recognize of sensory images and the system to construct individual purposeful movements guided by and adapted to the synthesized model of the environment.

• Vladimir A. Konyshev, Head of Neurobotics.

Development of an animatronic head with elements of communicative and emotional behavior.

• Professor Alexander A. Frolov, PhD. in Biology, Chief of the Mathematical Neurobiology Learning Lab at the Institute of Higher Nervous Activity and Neurophysiology (Russian Academy of Sciences).

Creation of a humanoid robot and a neural network controller that ensures the sustainable management of robot's movements through both direct communication and feedback.

• Professor Vladimir G. Yakhno, Ph.D., Head of the Auto-wave Processes Group, Laboratory Chief at the Institute of Applied Physics (Russian Academy of Sciences).

Creation of a system to analyse images from the set of sensors and select the objects of the specified class, matching their description in the operator’s language. Oriented for “parsing debris”.

3. The system of purposeful autonomous behavior (including a system to simulate emotional behavior), implemented in a fully autonomous mode to meet general goals set by a human operator.

• Professor Vladimir G. Yakhno Ph.D., Head of the Auto-wave Processes Group, Laboratory Chief at the Institute of Applied Physics (Russian Academy of Sciences).

The system to analyse images from the set of sensors and select the objects of the specified class, matching their description in the operator’s language. Oriented for “parsing debris”.

• Lev A. Stankevich, Ph.D. in Engineering, Professor of Department of System Analysis.

Technical proposal for the development of artificial neural systems of a humanoid robot (avatar).

III. Development of external control system via BCI:

1. Control system via BCI, based on recognition of EEG indices that correspond to movements, detected from motor cortex.

• Vladimir A. Konyshev, Head of Neurobotics.

Development of an animatronic head with elements of communicative and emotional behavior.

2. Control system via BCI with biofeedback training.

• Professor Alexander Y. Kaplan, Ph.D. in Biology, Head of the Neurophysiology and Neural Interfaces Lab at the Russian State University Biology Department (MSU).

Theoretical and experimental development of prototyping solutions controlled by EEG manipulators and robotic devices.

3. System to provide the operator with sensory data from the robot.

• Professor Alexander Y. Kaplan, Ph.D. in Biology, Head of the Neurophysiology and Neural Interfaces Lab at the Russian State University Biology Department (MSU).

Theoretical and experimental development of prototyping solutions controlled by EEG manipulators and robotic devices.

Each project contains a a concrete development. Applications with detailed descriptions are contained in the the Appendix.