Abstract:“The U t M Dextrous Robotic Hand” was the topic of the Plenary Speech given by Stephen C. Jacobsen at the Asilomar Conference. Jacobsen uses a video display to illustrate some of the capabilities of the hand. The angular finger deftly secures and cracks an egg against the side of a cup, whipping its contents into a froth. On command, one of its joints rings a bell at the speed of 70 times per second. This performance is only part of the repertoire of the “Utah/MIT dextrous hand,” a joint project of the University of Utah Center for Biomedical Design and the Artificial Intelligence Laboratory at the Massachusetts Institute of Technology. Under development since the early 1980s, 20 devices ultimately will be built to serve as test beds for artificial intelligence and artificial manipulation in laboratories across the U.S., says Dr. Jacobsen, Director of the Center for Biomedical Design and one of the system’s developers. Utah researchers have been involved in the more practical aspects of designing the hand and the lower level control systems that manage it. MIT scientists, delving into the abstract area of control theory, have combined computer science with mathematics to develop the computer and higher level control systems. Unlike industrial robots, which are automated, highly inflexible machines dedicated to performing specific tasks such as spot welding and painting, the dextrous hand is a high-performance device built to behave and perform like its biological counterpart. Newly designed tendons, made of KevlonDacron composites, operate a combined 19 joints in the hand, which also has multichannel touch-sensing capability. Now being designed is a tactile system of between 500 and 2000 sensors covering the entire palm of the hand that will tell a computer precisely where objects contact the hand and with what degree of force. The sensors, acting as a data-gathering system, will enable the machine to leani about its environment, to explore, iden@, and recognize objects around it, and to modify the hand’s behavior in response to that information. A computer, designed and assembled at MIT, controls a system that is roughly equal to five Apple Macintoshes, all of which interlace. For instance, there is one computer for each finger, and an integrated computer supervises the whole system. “Robotics bumps up against every technical limit you can possibly think of, from materials strength to energy conservation,” says Jacobsen. “Our objectives were to sacrifice everything but performance.” The researchers therefore built a system that is both complicated and consumptive of power. A compressor, bigger than a desk, generates 5 hp and must run constantly just to keep the hand at rest. “But its performance is extraordinary,“ Jacobsen adds. “It has roughly human strength, yet it’s quite a bit faster. Even more important than speed and strength is an elusive quality related to mechanical impedance.” “Good performance in people isn’t simply because of good control; it’s because the intrinsic behavior of the elements is right,” Jacobsen says. “You interact with objects as much as they interact with you. People do not pick up objects by positioning every finger like a robot would. So if a system doesn’t have these kinds of impedance properties, soft graceful motions and interactive kinds of behavior aren’t possible,” he says. “But if you do it intrinsically, relying on the physics of the actuators, then it happens automatically.” “Everything I’ve always looked at in biology has been that way. The physics are right first, then the control is simply used to modulate. Engineers tend to forget the intrinsics and try to control a device with smarts. It’s a philosophy that’s changing,’‘ Jacobsen observes. “Graceful behavior of muscle and bone components in biological systems is not a superfluous quality added later by way of neutral control systems to enhance aesthetics,’’ Jacobsen says. “Rather, it is a byproduct of appropriate component behavior. We considered these principles when designing the dextrous hand.“ The first application for the hand will be as a research tool to investigate issues related to artificial manipulation where a high degree of dexterity is needed. Later, the system will be simplified for use as an industrial robot, a teleoperation system (in which an operator’s hand movement activates the robot hand), and perhaps in prosthetics.

Seroprevalence of Toxocara canis antibodies in humans in northern Jordan

Clinical and genetic delineation of autosomal recessive and dominant ACTL6B-related developmental brain disorders

The clinical and molecular spectrum of the KDM6B-related neurodevelopmental disorder

Dextrous robotic hand

Biallelic ZBTB11 Variants: A Neurodevelopmental Condition with Progressive Complex Movement Disorders

De novo variants in ATXN7L3 lead to developmental delay, hypotonia and distinctive facial features

Mutations in EPG5 are associated with a wide spectrum of neurodevelopmental and neurodegenerative disorders

Biallelic pathogenic variants in TRMT1 disrupt tRNA modification and induce a syndromic neurodevelopmental disorder

Mutation in ZDHHC15 Leads to Hypotonic Cerebral Palsy, Autism, Epilepsy, and Intellectual Disability

Low endotoxic activities of synthetic Salmonella-type lipid A with an additional acyloxyacyl group on the 2-amino group of beta (1-6) glucosamine disaccharide 1,4'-bisphosphate

Identification of 11 Potentially Relevant Gene Mutations Involved in Growth Retardation, Intellectual Disability, Joint Contracture, and Hepatopathy

Clinical and genetic characterization of patients segregating variants in KPTN, MINPP1, NGLY1, AP4B1, and SON underlying neurodevelopmental disorders: Genetic and phenotypic expansion

Biallelic CRELD1 variants cause a multisystem syndrome, including neurodevelopmental phenotypes, cardiac dysrhythmias, and frequent infections

Biallelic loss-of-function variants of ZFTRAF1 cause neurodevelopmental disorder with microcephaly and hypotonia

De novo MCM6 variants in neurodevelopmental disorders: a recognizable phenotype related to zinc binding residues

Pathogenic Variants in TNRC6B Cause a Genetic Disorder Characterised by Developmental Delay/intellectual Disability and a Spectrum of Neurobehavioural Phenotypes Including Autism and ADHD.

Mechanism of KMT5B haploinsufficiency in neurodevelopment in humans and mice

Clinical and neurogenetic characterisation of autosomal recessive RBL2-associated progressive neurodevelopmental disorder

[Intra-arterial infusion of adriamycin in the treatment of locally advanced breast cancer].

Bi-allelic loss-of-function variants in BCAS3 cause a syndromic neurodevelopmental disorder

Unveiling the crucial neuronal role of the proteasomal ATPase subunit gene in neurodevelopmental proteasomopathies