Taking the Man out of Unmanned

header-robotics-unmanned

In the world of robotic science fiction, fantasies run wild. But in the reality of robotics development – whether they’re flying, swimming, or cruising through space – all unmanned systems still require a person to be in control at varying levels to operate.

In 1918, Lockheed Martin developed the Kettering Aerial Torpedo, the world’s first unmanned aerial vehicle. Technology has come a long way since then, adding advanced sensors, computing power, and communications capability to support increasingly autonomous – or independent – operation, but researchers still haven’t cracked the code on how to create fully autonomous system.

When we look to nature, we see the crow as an unassuming model of remarkable intelligence, having the ability to sense its environment and adapt to make informed decisions and keep itself safe[1]. An example of this can be seen in the way the birds have learned to crack tough nuts – they drop them onto a trafficked street, wait for a car to crush them, and then retrieve the opened nut when the coast is clear. 

“Today’s unmanned systems are not yet able to independently adapt to their environments like the crow, but they have proven their value,” said Dr. Al Romig, vice president of Engineering and Advanced Systems at Lockheed Martin’s Skunk Works®. “We are working to take more of the ‘man’ out of unmanned and provide our systems with better intelligence.”

Look Mom, No Hands
While unmanned systems are not fully autonomous, they aren’t operated like remote-controlled toy planes or cars either. Behind every one of our unmanned systems you will find Command and Control (C2) hardware and software helping automate navigational controls based on a pre-planned mission, and allowing the operator to focus on controlling and monitoring on-board sensors for faster decision making.

For example, our unmanned aerial systems (UAS) rely on a common control system (CCS) design philosophy.  With a CCS, the operator pre-programs the flight plan and uses sensor data to change the mission in real-time to re-task the plane. And whether it’s fixed-base or a mobile operation, a single CCS enables the operator to integrate and share data among multiple systems.

In a recent simulation, Lockheed Martin demonstrated exactly how a single operator can control, task and re-task multiple UASs simultaneously using an integrated C2 system.  By linking multiple platforms, sensors, and information collected into one enterprise view of the battle space, the common control system increases the speed of decision making.

Greater Autonomy Equals Greater Benefits
The reality is that even with advanced C2 solutions, today’s unmanned systems operators must endure long missions and analysts must review loads of data. While the technology continues to mature, turning autonomous technologies into a fielded reality will be no easy feat. Increasing autonomy will require increasing computational power and, possibly, a very different way of achieving that processing. 

Stephen Justice, director of Advanced Systems Development at the Skunk Works said,  “Our initial attempts at facilitating machine learning will doubtlessly mimic how we believe our brains learn – they will collect data, separate the relevant from the irrelevant, and then store that data as known trends.”

He explained that as new situations are introduced, the data will be compared to what’s known in order to draw a conclusion as to a proper response. For incoming information that has no known reference, that new data will need to be stored and trends developed to represent that data for use in future comparisons.

Until systems are as smart as a crow, Lockheed Martin will continue to leverage its experience to develop common control system architectures to make operations as efficient, safe and intuitive as possible.

“It is not enough to simply develop good ideas and new technologies, but rather these technologies must be tailored to be practical, useful, and fit the scope of the desired functionality in order to get a developed system from a lab environment into the battlefield,” added Romig.

1. Shettleworth, Sara J. (2010). Cognition, Evolution, and Behavior. Oxford University Press. pp. 3–4. ISBN 978-0-19-531984-2.

December 20, 2013

100th-bottomNavBar
   
IWP-logo
highlights
  • Whether they’re flying, swimming, or cruising through space, unmanned systems operate with some degree of autonomy. Yet they all require a person to be in control at varying levels.
  • Today’s unmanned systems are not yet able to independently adapt to their environments like the crow, but they have proven their value.
  • It is not enough to simply develop good ideas and new technologies, but rather these technologies must be tailored to be practical, useful, and fit the scope of the desired functionality in order to get a developed system from a lab environment into the battlefield

Speaking of the Future: Robotics


fish-pen-695

It may sound fishy, but Lockheed Martin’s innovation remote operating technology helps operate a mobile fish pen. Named a Top 25 Invention in 2012 by TIME Magazine, the mobile fish pen is a drifting fish cage that is highly automated and operates by integrating satellite communications, remote sensing data feeds, robotics, motor controls and LM’s C2 and situational awareness software.


Learn more about Data Analytics