October 23, 2008: HRL Laboratories LLC will begin research to develop electronics that simulate the cognitive capabilities and efficiencies of the biological brain as part of the Defense Advanced Research Project Agency’s (DARPA) SyNAPSE, or Systems of Neuromorphic Adaptive Plastic Scalable Electronics, program. HRL will lead a group of industry and university research laboratories with expertise in core areas of neuro and cognitive science in the groundbreaking research.

The goal of the SyNAPSE program is to bridge biology and electronics and establish a new paradigm for creating more intelligent machines that can interact with, react to, and actually learn from their environments.

“Two things are enabling us to establish this new paradigm,” said Narayan Srinivasa, SyNAPSE program manager and senior scientist in HRL’s Information and System Sciences Office. “First is the advancement in microelectronics, which provides us greater density and speed at much lower power consumption than ever before. The other is a far better understanding of neuroscience and how the brain functions at the synaptic, neuronal, network and system levels.”

The HRL team’s ultimate goal is to build a low-power, compact electronic chip combining a novel analog circuit design and a neuroscience-inspired architecture that can address a wide range of cognitive abilities — perception, planning, decision making, and motor control. In the initial two phases of the SyNAPSE program, the team will translate the neuronal and synaptic functions of the biological cortex into similar microelectronic functions.

The first nine-month phase of the program will focus on designing, fabricating, and characterizing synaptic and neural elements and combining them into a high-density, interconnecting microelectronic “fabric,” which will be incorporated into a more complex system-level fabric design. The team will also develop simulation methods that integrate neural brain models with a virtual environment in which to test a sequence of tasks, including visual perception, planning, navigation and various types of learning.

In the following 15-month phase, HRL will combine the synaptic and neural elements to fabricate and demonstrate “cortical microcircuits” that can model various lower-level brain functions and actually “learn” by interacting with the environment. During this phase, extensive simulations and further testing will occur to verify the “behavioral competence” of the brain models.

“The follow-on phases of the project will create a technology that has the neurological capabilities of a brain with the density of approximately 100 million neurons and one trillion synapses,” Srinivasa said.