Creating and Preventing Strategic Surprise
Our Work Banner
Robotics Challenge

Robots Learn to Perform Tasks by “Watching” YouTube Videos

University of Maryland computer scientist Yiannis Aloimonos (center) is developing robotic systems able to visually recognize objects and generate new behavior based on those observations. DARPA is funding this research through its Mathematics of Sensing, Exploitation and Execution (MSEE) program. (University of Maryland Photo)
Robots can learn to recognize objects and patterns fairly well, but to interpret and be able to act on visual input is much more difficult. Researchers at the University of Maryland, funded by DARPA’s Mathematics of Sensing, Exploitation and Execution (MSEE) program, recently developed a system that enabled robots to process visual data from a series of “how to” cooking videos on YouTube. Based on what was shown on a video, robots were able to recognize, grab and manipulate the correct kitchen utensil or object and perform the demonstrated task with high accuracy—without additional human input or programming. Article 
University of Maryland computer scientist Yiannis Aloimonos (center) is developing robotic systems able to visually recognize objects and generate new behavior based on those observations. DARPA is funding this research through its Mathematics of Sensing, Exploitation and Execution (MSEE) program. (University of Maryland Photo)
Code 144

Establishing the CODE for Unmanned Aircraft to Fly as Collaborative Teams

DARPA’s Collaborative Operations in Denied Environment (CODE) program aims to develop algorithms and software that would extend the mission capabilities of existing unmanned aircraft systems (UAS) well beyond the current state of the art, with the goal of improving U.S. forces’ ability to conduct operations in denied or contested airspace. CODE would enable mixed teams of unmanned aircraft to find targets and engage them as appropriate under established rules of engagement, leverage nearby CODE-enabled systems with minimal supervision, and adapt to situations due to attrition of friendly forces or the emergence of unanticipated threats—all under the command of a single human mission supervisor. CODE envisions improvements that would help transform UAS operations from requiring multiple people to operate a single UAS to having one person able to oversee six or more unmanned vehicles simultaneously. 

The U.S. military’s investments in unmanned aircraft systems (UAS) have proven invaluable for missions from intelligence, surveillance and reconnaissance (ISR) to tactical strike. Most of the current systems, however, require constant control by a dedicated pilot and sensor operator as well as a large number of analysts, all via telemetry. These requirements severely limit the scalability and cost-effectiveness of UAS operations and pose operational challenges in dynamic, long-distance engagements with highly mobile targets in contested electromagnetic environments. Article

Atlas Unplugged 2 - 144

Upgraded Atlas Robot to Go Wireless as the Stakes Are Raised for the DARPA Robotics Challenge Finals

The DARPA Robotics Challenge teams using the Atlas robot met in Waltham, Mass., on January 12, 2015, to learn about upgrades to the robot. (DARPA image courtesy of Worcester Polytechnic Institute) 

A total of $3.5 million in prizes will now be awarded to the top three finishers in the DARPA Robotics Challenge (DRC), the final event of which will be held June 5-6, 2015, at Fairplex in Pomona, Calif. The new prize structure was created in recognition of both the significant progress already demonstrated by teams toward development of human-supervised robot technology for disaster response and the increased number of teams planning to compete in the Finals, including those funded by the European Union and the governments of Japan and South Korea. Aside from the previously announced $2 million grand prize, DARPA plans to award $1 million to the runner-up and $500,000 to the third-place team. DARPA expects at least twenty teams to compete in the DRC Finals.  Article

Equips 144

Minimizing Uncertainty in Designing Complex Military Systems

DARPA’s Enabling Quantification of Uncertainty in Physical Systems (EQUiPS) program aims to develop mathematical tools and methods to efficiently quantify, propagate and manage multiple sources of uncertainty. 

Uncertainty is sometimes unavoidable. But in the world of scientific computing and engineering, at least, what’s worse than uncertainty is being uncertain about how uncertain one is. Article

Matrix 144

Developing New Materials For Energy Transduction

DARPA’s Materials for Transduction (MATRIX) program seeks new classes of materials that convert energy from one form into another that can be demonstrated directly in defense applications and devices. The program aims to advance innovative modeling and simulation tools that engineers can use to design systems that take advantage of these new materials and their energy-transforming properties. 

Transduction involving the conversion of energy from one form into another is common in many military and space devices, such as communications antennas (radio waves to electrical signals), thermoelectric generators (heat to electricity) and electric motors (electromagnetic to kinetic energy). Research efforts to develop new transductional materials, however, have largely been limited to laboratory demonstrations and haven’t always resulted in new capabilities or significant size, weight, and power (SWAP) reduction for military devices and systems. Article 

2014 Video logo

The Top 10 Most Popular DARPA Videos of 2014

Shield 

The DARPA YouTube channel receives millions of visits each year. In 2014, we shared information about new efforts and announced milestones reached in our existing programs. A full list of videos is available at http://ow.ly/G88w2. A list of the top 10 most popular DARPA web features of 2014 is available at http://go.usa.gov/e8t3. Article

Share this page: