Revolutionizing Prosthetics

Overview

Begun in 2006, the Defense Advanced Research Project Agency’s (DARPA) Revolutionizing Prosthetics program set out to expand prosthetic arm options for today’s wounded warriors. The program funded two teams to create advanced anthropomorphic mechanical arms and control systems: DEKA Research and Development Corporation to get an arm control system to market quickly, and the other—with APL as the system integrator and lead—to produce a fully neurally integrated upper-extremity prosthesis with appropriate documentation for clinical trials and manufacturing transition. APL created a modular architecture and extensible platform that provided a framework for future developments by us or others.

A prosthesis requires the following characteristics in order to reflect the properties of a biological limb:

sensors for touch, temperature, vibration, and proprioception (the ability to sense the position of the arm and hand relative to other parts of the body)
power that will allow extended use
mechanical components that will provide strength and environmental tolerance (to heat, cold, water, humidity, dust, etc.)

With this new prosthetic, an upper-extremity amputee would be able to feel and manipulate objects just like a person with a native hand.

Integrated Technologies

Virtual Integration Environment (January 2006-present)

Underlying the research, design, and development during all phases of the Revolutionizing Prosthetics program is the Virtual Integration Environment (VIE). The VIE is modular and configurable to support various limb models and algorithms in order to facilitate highly dexterous control of an upper-limb prosthesis. It serves as a complete limb system simulation environment to support neural integration. The VIE is used to visualize and monitor performance of various design approaches, pilot neural signal analysis algorithms, and simulate emerging mechatronic elements. It provides a framework for synchronizing research and efforts across many geographically distributed contributors in order to turn neural research into engineering reality. From the clinical and patient perspective, the VIE is used to train end users to control real or virtual neuroprosthetic devices, and configure and customize clinical and take-home devices.

VIE development is continuing with enhancements for a simplified, platform-independent communication interface to the MPL and an improved three-dimensional MPL simulation environment with scriptable scenarios.

Neural Research (January 2006-present)

Since the beginning of the Revolutionizing Prosthetics program, researchers have advanced the state of the science in neuroprosthetic control. We are currently focused on technology demonstrations of cortical control of the MPL by spinal cord injury patients. (Note: The National Institutes of Health [NIH] is also supporting other brain–computer interface (BCI) investigators in demonstrating that their approaches might be effective in controlling the MPL.) The neuroscientific goals are to evaluate the efficacy of cortical microstimulation for tactile and proprioceptive feedback in humans; to develop active and wireless implantable microelectrodes for stimulating, recording, and demonstrating the viability of chronic implantation; and to evaluate the feasibility of bilateral closed-loop cortical control. Experiments for the first-ever neuroprosthetic system controlled with signals from the posterior parietal cortex and the first-ever demonstration of closed-loop cortical control of a dexterous neuroprosthesis using recordings from the motor cortex and parietal reach region and stimulation in the primary somatosensory cortex are planned. Additionally, we will identify commercial transition partners for neuroprosthetic technologies.

The research breakthroughs resulting from these activities will provide prosthetic limb control options not only to amputees but also to patients with an inability to control their native limbs as a result of stroke or degenerative conditions such as amyotrophic lateral sclerosis (ALS).

Modular Prosthetic Limb (December 2009-present)

The first version of the Modular Prosthetic Limb (MPL v1.0) was completed in December 2009. It featured anthropomorphic form factor and appearance, human-like strength and dexterity, high-resolution tactile and position sensing, and a cortical neural interface for intuitive and natural closed-loop control. Made of lightweight carbon fiber and high-strength alloys, the arm has 25 degrees of freedom, or joint motions (the human arm has about 30). This limb is modular and configurable to the level of a patient’s injury. MPL v2.0 was completed in December 2010, and our current efforts focus on refinement of the MPL system technologies by incremental design improvements in order to improve reliability, increase simplicity, leverage commonality among subsystems, maximize performance, and evolve software and controls algorithms. MPL systems are being distributed to clinical partners for use in neural-based research and development of cortical encoding and decoding strategies.

Proto 2 (August 2007)

During development of Proto 1, the Revolutionizing Prosthetics team engaged in concurrent research, analysis of alternatives, design, and development efforts for design of a limb including: biomimetic operation including all degrees of freedom, speed, dexterity, and force; electromechanical actuation mechanisms; communications, signal analysis, and control strategies; and comfort and appearance (socket and cosmesis). The mechatronic culmination of these efforts was completed in August 2007. Two versions of Proto 2 were designed—one with an intrinsic hand using off-the-shelf brushless direct current motors, and one with an extrinsic hand using a cobotic device in the forearm with a single motor to control the wrist and hand. There are about 25 microprocessors (computer chips) and 80 tactile sensors built into the fingers, fingertips, hand, wrist, and elbow. The Proto 2 limbs incorporated significant advances in many technologies, leading to the ultimate design of the MPL.

Proto 1 (December 2006)

Prototype 1 (Proto 1) has eight degrees of freedom—a level of control far beyond the current state of the art for prosthetic limbs. As a rapid-development prototype, Proto 1 utilized many commercial off-the-shelf components with the purpose of supporting neural integration research and serving as a test bed for evaluation of haptic feedback and indirect sensory perception approaches. Additionally, Proto 1 was used to demonstrate advanced prosthetic function with classification algorithms for noninvasive and low-invasive control devices. Proto 1 was fitted for the beginning of Revolutionizing Prosthetics clinical evaluations using surface electromyographic control with targeted muscle reinnervation (TMR) patients at the Rehabilitation Institute of Chicago in January 2007.

Timeline

Major Events

2019	First human (Revolutionizing Prosthetics Participant 4) is implanted with six microelectrode arrays, two in the primary motor cortex (M1) hand area and two in the primary sensory cortex (S1) hand area of the left hemisphere of the participant (the dominant one, because the participant is right-handed), and one in M1 and one in S1 of the right (nondominant) side.
August 2017	FDA approves implantation of microelectrode arrays in the motor cortices and somatosensory cortices on both hemispheres in a human candidate.
2016–2017	Revolutionizing Prosthetics Participant 3 demonstrates control over multiple virtual aircraft in a pilot simulator.
2016–2017	Revolutionizing Prosthetics Participant 3 is able to perceive forces placed on sensors in each of the fingertips of the MPL through stimulation of the somatosensory cortex.
2015	First human (Revolutionizing Prosthetics Participant 3) is implanted with four microelectrode arrays, two in the motor cortex and two in the sensory cortex of the left side of the participant (dominant, because the participant is right-handed).
2014	First human (Revolutionizing Prosthetics Participant 2) demonstrates non-anthropomorphic control of a flight simulator.
May 2014	MPL v3.0 is completed.
2013	FDA approves implantation of microelectrode arrays in somatosensory (S1) cortex for recording and imparting electrical stimulation.
April 2013	First human is implanted with two microelectrode arrays in parietal cortex (AIP and BA5), with no device-related adverse events to date.
December 2012	“Breakthrough: Robotic limbs moved by the mind” is featured on CBS’ 60 Minutes.
December 2012	Clinical trial results are published in The Lancet.
September 2011	The MPL is controlled by a patient using ECoG during a UPMC and University of Pittsburgh experiment.
February 2011	First human is implanted with two microelectrode arrays in primary motor cortex (M1) hand area, with no device-related adverse events to date.
December 2010	MPL v2.0 is completed.
September 2010	Phase 3 Kickoff meeting for APL Revolutionizing Prosthetics team.
December 2009	Version 1 of the Modular Prosthetic Limb (MPL v1.0) was completed with 17 degrees of freedom.
April 2008	Phase 2 Kickoff meeting for APL Revolutionizing Prosthetics team.
August 2007	Two versions of Proto 2 were completed with 22 degrees of freedom. The extrinsically actuated hand employed a cobot in the forearm driving the hand, wrist, and radial rotator. The intrinsically actuated hand was motor driven.
January 2007	Prototype 1 (Proto 1), with 8 degrees of freedom, began preclinical evaluations using surface myoelectric control. The Virtual Integration Environment (VIE) was used for clinician interface and patient training.
January 2006	Phase 1 Kickoff meeting for APL Revolutionizing Prosthetics team.

Program Background

DARPA contracted more than $107 million for the APL-led Revolutionizing Prosthetics program to create a fully functional upper limb that responded to direct neural control. The groundbreaking work done under this program capitalized on previous DARPA investments in neuroscience, robotics, sensors, power systems, and actuation. In particular, this program built on DARPA’s Human Assisted Neural Devices program, which decoded the brain’s motor signals with such fidelity that movements of a robotic arm could be controlled entirely by direct brain control.

The following items illustrate DARPA’s rationale, goals, and commitment to Revolutionizing Prosthetics and associated programs.

Partners Build High-Tech Hands for Wounded Warriors, by Cheryl Pellerin (American Forces Press Service, December 15, 2010)
DARPA’s Bio-Revolution, by Jonathan Beard (DARPA: 50 Years of Bridging the Gap, April 2008, pp. 158–160)
DARPA Revolutionizes Prosthetics: How and Why?, by Judith Philipps Otto (The O&P Edge, November 2007)
DARPA’s Cutting-Edge Programs Revolutionize Prosthetics, by Donna Miles (American Forces Press Service, February 8, 2006)

Partners

The Revolutionizing Prosthetics team brought together the most respected scientific researchers in their fields, as well as commercial leaders from the prosthetics industry. Often referred to as the “Manhattan Project” for prosthetics, the project was shaped by a multidisciplinary team that included experts in signal processing and pattern recognition, electrodes for brain implants, mechatronics, neuroscience, electrical engineering, cognitive science, signal processing, battery design, nanotechnology, and even behavioral science.

Revolutionizing Prosthetics

Overview

Virtual Integration Environment (January 2006-present)

Neural Research (January 2006-present)

Modular Prosthetic Limb (December 2009-present)

Proto 2 (August 2007)

Proto 1 (December 2006)

Major Events

Sponsors

Program Background

Partners

Contact Information

Media Relations

Paulette Campbell

Technology Transfer

Heather Curran

Other Inquiries

Revolutionizing Prosthetics Team