Speaking to a trade conference in January 2007, Microsoft co-founder, Bill Gates, argued that the next big technological horizon will be in robotics and projected that the field is currently in the same position as the computer industry was in the mid-1970s on the verge of the Home PC explosion. If that is the case, predicting the future of robotics will be as difficult as it would have been to imagine a mind-controlled humanoid robot 30 years ago. This may be cause for skepticism among some, and may remind others from a previous generation of the failed promises of robots to live up to expectations. While developing the sophisticated “brains” of the robot has been a long-anticipated challenge to overcome, experts are beginning to project that in the next few decades artificial intelligence will have attained the ability to mimic human cognitive processing and to perform at hundreds of billions of calculations per second.
In order to achieve the levels of sophistication required to create synthetic parts for the next generation of humanoid robots, new disciplines of study and training are necessary. A major area that holds promise for development into humanoid technologies is neurobioengineering. The field itself is not new, but was founded in 1995 by professor Massimo Grattarola (now deceased) of the Biophysics and Electrical Engineering Department (DIBE) at the University of Genova, in Genova, Italy. Grattarola’s efforts were related to his design of an undergraduate and graduate program named neurobioengineering, at the university where he worked.
Grattarola’s program attested to his multidisciplinary ingenuity, as it was—according to one source—“designed to amalgamate anthropomorphic robotics, artificial intelligence, bioelectronics, electrical engineering, molecular biology, physics, and medicine, into a single program with the aim of developing advanced bio-compatible neuro prosthetic implants” (i.e., artificial limbs, central and peripheral nervous system implants, directional neural grafting). The multidisciplinary nature of neurobioengineering implies a significant cross section of core competencies between electronics, biological systems, and neuroscience. In particular, a deep understanding of the electrical nature of the brain is critical for engineers who work in this field and who must be equipped with the ability to reverse engineer the human brain for machine applications. In order to better understand the neural circuitry of the brain, a brief overview of the function of neurons and synapses is helpful.
Neurons, or nerve cells are comprised of three basic components: the cell body (soma), axons, and dendrons; the cell body contains the nucleus and other vital components of the cell; the axon is a cable-like projection that carries the action potential away from the cell body along the length of the neuron; and dendrons are branch-like projections situated at the ends of the neuron that enable it to communicate with other neurons. The process of synaptic transmission involves a highly complex interchange between the electrical charge, the axon terminal, the pre- and post-synaptic cleft, and the neurotransmitters that cross this cleft through chemically charged ions. Neurotransmission occurs as the action potential, or electrical impulse, travels to the synaptic terminal (the presynaptic ending), situated at the end of the axon. This impulse will trigger the migration of synaptic vesicles containing neurotransmitter molecules toward the presynaptic membrane. This vesicle membrane and the neurostransmitters will fuse with the presynaptic membrane at the end of the axon, releasing the neurotransmitters into the synaptic cleft, which then bind with ion channels that get transmitted as an electric charge to the dendrite of the postsynaptic neuron. Once the electric potential is transferred to the dendrite (or dendron branch) it continues on to the next neuron, thus propagating the signal to additional neurons in a complex series of electo-chemical interchanges.
The “Holy Grail” of robotics research has been about replicating this complex function of the human brain into a synthetic format that offers a working solution for the development of sophisticated humanoids. There are two distinct techniques to how this may be accomplished; 1) replicating the brain through construction of scalable synthetic structures. One major project currently under way at USC, for example, is working on a “synthetic cortex” that relies upon neurons built from carbon nanotubes, and fabricated together in such a manner that will attempt to replicate aspects of electro-chemical ‘synaptic transmission’ in the human brain; or 2) the creation of a logic electronic circuits from rat cell neurons grown in a geometric design. This is a project that met with some well-broadcast success in the application to robots, and so the possibility of creating innovative hybrid circuits that are part organic and part mechanical is a present reality that must be factored into the future of humanoid robotics development.
The future of robotics will look significantly different than anything propagated to this point through the traditional disciplines of computer science, mechanical engineering, and electrical engineering. Going forward, the development and advancement of scalable and open-domain resources and techniques will rely on various sets of new multi-disciplinary collaborations between the realms of artificial intelligence, bioelectronics, nanotechnology, molecular biology, physics, biomimetics, and neuropomorphic systems. As a result of these explorations in human robot interaction, especially the interface between human and humanoid technologies, one can begin to envision the exciting advances in store for the future of humanoid robotics!
Tags: bioelectronics, future of robotics, human robot interaction, humanoid robotics, humanoid robots, neurobioengineering, neuronal circuits, neuroscience
Cool! Thanks for sharing this!