Tech Talk: Enhanced Perception for Autonomous Vehicles [Transcript]

By John Ohab

The following is a transcript from the Office of Naval Research’s (ONR) online Tech Talk series.

Keith Hammack, deputy manager for ONR’s Maneuver Thrust research area, answered questions from audiences on Facebook and Twitter on Sept. 22.

Previously deployed as a combat company commander with the U.S. Army, Hammack (right) today manages research programs in ground vehicle autonomy, fuel efficiency, mobile power, mobility, and survivability for the Marine Corps, Naval Expeditionary Combat Command and Naval Special Warfare Command.

Editors’ Note: The following transcript includes questions submitted earlier on Facebook and Twitter, and e-mailed directly to ONR.

Event Transcript

ONR: Welcome to our online dialogue with Keith Hammack, deputy director for ONR’s Maneuver Thrust. We will take audience questions on a first-come, first-served basis until 11:45 a.m. Responses will be posted in this conversational field. Comments should appear quickly, but for up-to-the-minute postings, you may need to refresh your browser window periodically during the event.

ONR: Our first question comes from Bob Finkelstein of Robotic Technology Inc. via e-mail.

ONR: Bob asks if ONR would be interested in a project where the focus is on the development and application of an autonomous, intelligent control system architecture to achieve the desired perception, independent of specific sensors.

Keith: Bob, we have met a couple times, including at the Intelligent Vehicle Technology Transfer (IVTT) meeting that you organized at the National Institute of Standards and Technology in 2009. We would absolutely be interested in learning more about the architecture you are referencing. I wonder if it is similar or the same as the 4D/RCS that you had mentioned to me in the past?

Keith: To add, if you would like to talk in more detail, come out to Modern Day Marine at Quantico Marine Corps Base, Sept. 28 – Sept 30. We’ll be located in Tent C, where we will have our largest ONR display ever highlighting ONR’s autonomy work and science and technology in support of U.S. Marine Corps ground vehicle modernization.

ONR: Next question comes from Ruben Potts, who asks what type of sensors are good for autonomous vehicles?

Keith: Ruben, many current ground vehicle efforts use LIDAR (light detection and ranging) to sense their surroundings and whether objects are in front of or near them. The challenge is classifying the object (e.g., rock, bush). We are investigating the addition of electro-optic sensors, or cameras, to these platforms to augment or replace the LIDAR sensors.

Keith: To add, cameras, much like our eyes, provide a larger amount of data than most other sensors. Since equipping small vehicles with supercomputers is neither feasible nor affordable, we have to couple these cameras with smart algorithms to glean the data we need efficiently for near real-time decision-making capability. Any other small, low-cost rugged sensor that could help provide detection and classification would be considered also.

Chris Harding: Significantly improved thermal management and power density electro-mechanical power supplies and propulsion systems for small unmanned ground vehicles, robots, and exoskeletons.

■Has the DoD considered the use of micro-heat exchangers and chemical reactors, Oregon State University’s Dr. Jovanovic, to allow faster dispersion of heat from a system? This may allow heat to be quickly dispersed through a medium with a high thermal conductivity. If a heat signature is a concern, the immediate dispersal may reduce the effect of radiative heat and reduce overall temperature through a means similar to counter-current flow processes. The “specific surface area”, which is the relationship of volume to surface area is quite beneficial in these systems and allows larger fluxes of energy. Although fouling is a concern, it is a promising area of research that could “force” cleaner burning systems as well.

■Also, UWB is a low power consumption RF device, which translates to reduced energy consumption and heat production, and I am sure you are using this as a resource. Have you considered the use of waveguides that have been “doped” and specifically designed for measuring temperatures? The waveguides could be coupled with the microscale reactors and heat exchanger units to create process control for such delicate systems. As you know the physical or chemical properties of designed waveguides, fiber optics, can be designed to change upon changes in temperature, which must be similar to a vapor deposition process. Also, they can be designed to detect pH changes or concentrations as well.

■Actually, I just looked it up and it is called OVD instead of PVD or CVD. http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=1456097

Chris Harding: Oh yeah! I wanted to tell Mr. Hammack and all others thanks for their service!

Chris Harding: As a side note, I believe your scientist, technicians, engineers, and any other contributors would enjoy the following! I found it tonight because I wanted to know more about UWB, and I wouldn’t be surprised if someone already knows of the story. Search Google for: “On Shannon and “Shannon’s formula” and a PDF becomes available.

ONR: Chris Harding posted a multi-pronged question which we’ll tackle in several posts.

Keith: In answer to your earlier post, ONR’s Sea Warfare and Weapons Department has an Energy Conversion Discovery and Invention (D&I) Program that invests in long-term basic research to uncover new, more efficient energy conversion mechanisms as well as new tools for understanding the physics behind emerging energy conversion approaches.

Keith: To add, in the near term, it is likely that this system will have to be powered by an internal combustion engine similar to the DARPA “Big Dog” program. (See: http://www.youtube.com/results?search_query=big+dog&aq=f). The weight and duration of electrical power sources currently can’t support the mission needs of the Marines. We hope that great ideas like yours and our D&I portfolio will solve this issue over the coming years.

Chris Harding: @About LIDAR: When considering coatings, the “absorption” effects of various metals, polymers, causes major changes in the refractive index of the medium and can significantly affect information and may even cause diffusion of light in one …particular direction. Some materials can be designed to be more sensitive to different wavelengths. Is this similar to the optical devices being used?  I learned the latter in thin films and it’s a bit rusty!

Keith: Finally, the Sea Warfare & Weapons Department has a number of efforts looking at single and two-phase convective heat transfer from microchannels and microjets, including efforts at Oregon State. We can’t go into much detail on thermal signatures and related topics in this environment. If you would like to come in and talk with us further, plan to join us at the 2010 ONR Naval Science & Technology Conference in Arlington, Va., Nov. 8-10, 2010.

Trenton Truggle: Mr. Hammack, for the CNR Challenge, what’s the desired trade-off between prototype and design documentation? If, for example, a particular sensing technology is in its infancy and a submitter believes they can significantly advance the technology, are proof-of-concept engineering prototypes the de-facto “best” way to back up the claim?

ONR: Next question from Trenton Truggle asks about the upcoming CNR Challenge. He wants to know the desired trade-off between prototype and design documentation? If, for example, a particular sensing technology is in its infancy and a submitter believes they can significantly advance the technology, are proof-of-concept engineering prototypes the de-facto “best” way to back up the claim?

Chris Harding: @BigDog: Amazing process control!

Keith: Trenton, at ONR we invest in basic and applied research and advanced technology development with technologies readiness levels from 1-6. This means we are interested in developing and maturing new and novel ideas to fill the gaps in technologies based on warfighter needs. Our ultimate customer is the warfighter.

Keith: To add, technology demonstrators are often the result of our research, at which time we work with the performer to transition to a Marine Corps or Navy program office, such as the Robotics Systems Joint Program Office.

Chris Harding: @Keith: Amazing weaponization of BigDog!

Keith: Chris, going back to your LIDAR question, ruggedization of these sensors is a key concern for military applications. In other words, we need to make sure that these systems are not only effective, but can withstand the elements (e.g., sandstorms) and environment (e.g., shock, vibration).

Keith: There are many different sensors with various coatings. If you’d like to talk about a specific idea that you have, we would be happy to hear it. Check out the Maneuver Thrust web link and keep watching for our broad agency announcements in fiscal year 2011.

ONR: Final words from Keith?

Keith: If you have an idea that you’ve been pondering, but you’re concerned that it’s not ready for prime time, we can help you determine how to get it there. Partnering with bright minds from academia, industry and government labs is what we do — it’s vital to ONR’s ability to successfully deliver capabilities to the warfighter.

Chris Harding: Keith, I have done some work with particle sizing and learned a bit about light distinction, laser diffraction, and a few other methods. During that time, I learned that light extinction will often cause about a 25%, on average, variation… in the “diameter” of a particle. Obviously, the same particle could cause a change within the population distribution if not “spherical” and it makes multiple passes.

■In my opinion, this is similar to the problem you face with bushes versus rocks.

■Also, I have heard some people were attempting to use UWB technology to be used as a “finder” of variable densities. Is this true?

ONR: Great! Thanks to everyone for participating in this week’s “Tech Talk” with ONR. Join us next at 11 a.m., EST, Wednesday, Sept. 29, for a conversation with Dr. Mike Traweek on what the Navy can learn from fish and their innate ability to process sound despite flow noise.

Chris Harding: ‎@Mr. Hammack: Please remember there are many Americans who appreciate the service of our military! I have a friend that flies the Osprey, and I attempt to thank him regularly! The military has some brilliant researchers! Many don’t know that much of our “technology” comes from you all! Thanks!

Tamra Temple: Thank you.

On the next Tech Talk, Dr. Michael Traweek, program officer at ONR’s Maritime Sensing Team, will discuss Flow Noise Mitigation by Fish. Join the discussion at 11am eastern on Wednesday, Sept 28.