Archive for category robots

Summer Reading List

Posted by jasonwong in consciousness, decision making, robots on June 3rd, 2009

It has been an insane few weeks with moving from Virginia to Rhode Island, and then travel to the Midwest for family events, which is why there have been no new posts. However, I have been doing a lot of reading since I have not had much Internet access recently either.

Wired for War

Wired for War by P.W. Singer is a book about the coming robotic revolution in warfare. The author is a young analyst at the Brookings Institute, which is a political think tank. I was pleasantly surprised that the author was not a gung-ho military man nor was he an anti-military hippie. He is, however, a geek. The first part of the book details all the current and up and coming robots used by the military, and there is a sense of awe as he describes these military machines that, for a science fiction geek like myself, are really cool.

The second part of the book is also a fascinating read, but not in the same way as the first. Here, Singer discusses the implications of the robotic revolution, and there are many. A section of the book is dedicated to traditional human factors, discussing usability issues and the like. The discussion of who should be held responsible if a flying drone fired a missle on the wrong target is especially chilling. If the pilot entered the right coordinates and the machine messed up, should the programmers and engineers be held responsible? In a military trial?

wiredforwar

There are also discussions of how drone pilots live in Nevada, work a shift where they kill enemy combatants, then make it home in time for family dinner and a PTA meeting. It is a totally different experience from a Marine actually on the ground, but they are both soldiers, right?

All in all, the book is a great read. From whiz-bang technology to psychological and societal implications, this book is well-researched and considers all facets of the issues. The second half of the book could have used a more thorough editor, in my opinion, as I felt that the author repeated points across chapters and lost a natural flow of ideas because ideas where not contained in their own chapters.

If you can get past a mildly sloppy second half, however, the entire book is worth reading for anyone in the military (like myself as a civilian Human Factors Scientist) or anyone who has interests in military technology or even technologically-oriented science fiction buffs.

Blink

Blink by Malcom Gladwell has been out for a while, so I will keep my review short. The book is fantastic. Gladwell focuses on the idea of rapid cognition. He is careful to not call this intuition, which he feels is more emotional. Rapid cognition is more about near-instant unconscious thought. That split second decision that something is not quite right bit you are not sure what. Or what love at first sight is. Or how unconscious prejudices can affect our social interactions despite our conscious attitudes on race.

The topic itself is interesting, and so is Gladwell’s writing style. He weaves in appropriate and engaging stories that introduce his point and then often goes into psychological research the further illustrates his points. Finally, he talks about how these ideas have real-world implications for how to better live our lives.

In short, the book argues that our rapid cognition is a useful tool as long as we know where it can go wrong. Blink is the perfect example of an engaging pedagogy, and I definitely plan to incorporate this book the next time I teach my Cognitive Psychology class.

blink

Books Still on the List

I also plan on reading the human factors design classic The Design of Everyday Things. Donald Norman, the author, is a giant in the field (especially with regards to design) and this book is one of his earlier writings but remains one of the best. Considering I have my PhD in human factors, it is one of those books that I should read.

No Comments

The DoD needs cognitive psychologists!

Posted by jasonwong in attention, perception, robots on July 19th, 2008

SBIR call for “A psychologically inspired object recognition system”

The DoD has put out a call for proposals for the development of an object recognition system for computers that obeys psychological principles. Object recognition is obviously important for humans, and as more robots are being used in place of humans, they should also be able to identify objects to aid in mission success. The project description sounds like a short review of the object recognition literature, in fact (emphasis added by me):

Recognizing and identifying an object from a video input turns out to be a very difficult problem. The problem stems from the fact that a single object can be viewed from an infinite number of ways. By rotating, obscuring, or scaling a single object, one can create multiple representations of an object – which makes the problem of matching the object to a database of objects very difficult. The problem expands exponentially when objects that need to be identified have never been viewed before. Combine these limitations with the wide variety of objects which might be identified, and the problem becomes intractable. One solution is to study and understand how human beings recognize objects in the real world and duplicate that functionality in a series of algorithms. Recent research (Tarr and Bulthoff, 1995) has indicated that humans use not one algorithm, but multiple algorithms for the task of object recognition – depending on the object being recognized and the situation at hand. Specifically, research has shown that people use template based algorithms (i.e. similar to the database matching algorithms described earlier) in addition to Geon based (Beiderman, 1995) algorithms and feature based algorithms.

First of all, 1995 counts as recent research? Sounds like some DoD scientists need to attend the Vision Sciences conference. Secondly, it is satisfying to see that the DoD believes that understanding how the human mind works is a big step in implementing human-like cognition in artificial systems.

This is similar to the field of biorobotics, where the understanding of how natural organisms work (say, a dolphin) can be applied to machines (say, a submarine). This makes a lot of sense, actually. Biological organisms are highly evolved – nature has done the work of choosing what works best. By studying what works best, we can use those principles in designing our own machines. It seems like a new principle in engineering, but it makes a lot of sense.

No Comments

User interfaces in Iron Man

Posted by jasonwong in data visualization, eye movements, robots, training on May 3rd, 2008

First off: possible spoilers ahead!

Iron Man was an excellent movie. However, since this blog is not dedicated the movie reviews, I thought I’d discuss some of the user interface elements involved instead.

This video highlights two UI elements that were well thought-out. There are more I’d like to discuss, but I couldn’t find them. There are two clips: one is of a holographic prototyping interface and the other is the Iron Man suit user interface and flight interface:
Click to download (6.5 MB MP4 video)

Holographic Prototyping and Direct Manipulation on the Cheap:

The first clip ends at around 26 seconds and shows off Tony Stark’s (aka Iron Man) holographic prototype interface. The hologram is a sci-fi cliche, but its usefulness is immediately evident. Direct manuipulation has been discussed before, and this typically requires something physical to manipulate. These physical prototypes are expensive to fabricate, especially if multiple revisions are needed. In the clip, Stark has already built out the specs for this piece of his suit, and he’s able to add and subtract parts and accurately visualize the effects of the modifications without fabricating lots of physical prototypes. The coup de grace is when he is able to stick his arm inside of the hologram and test it out. It’s direct manipulation of a prototype without the expense. Because this kind of manipulation is so natural, very little cognitive effort is needed to use this interface.

Flight Suit Interface and Transfer of Training:

The second clip starts at 27 seconds and demonstrates part of the suit’s user interface and the flight interface as well. The flight interface is very similar to that of a fighter jet, which will make transfer-of-training easy from a fighter jet to an Iron Man suit. This will cut down on the need to train users of the Iron Man suit – if they can fly a fighter jet, they can fly this suit.

Main Suit User Interface Voice Commands:

What was most interesting in the clip was how the general UI was controlled. The heads-up display is directly in front of the user’s face, but the user cannot touch the display. Therefore, direct manipulation is out of the question. The suit does take voice commands, as shown in the video. This is an obvious choice, but it is slow to use. Imagine flying at some insane speed under high stress – do you want to have to yell out a command that takes several seconds to issue, then wait for a reply from the suit? Probably not a good idea. The closest thing to voice interaction in this world is the Microsoft Sync system. This system integrates bluetooth phones and MP3 players into Ford cars and is all voice controlled. When it works, the eyes stay on the road for longer and less attention is required to make a phone call or play music. But when it doesn’t work, the error correction is simply a mess. It takes a huge amount of effort and is bad for driving or flying.

This is a great review of the Sync system (see 1:45 for an example of a Sync error as how the user simply gives up):

Main Suit User Interface and Eyetracking:

Besides voice commands, the other control option is eyetracking. If the eyes are focused on something in the environment, a command can be issued to zoom in, take a picture, etc. Issuing that command, however, would have to be done using a voice command or a button-press. The eyes are needed to focus, so something other than the eyes must issue a command. This is not the ideal situation because it requires coordination of multiple systems – the eyes must remain focused on the target while another body part confirms the command. Overall, though, this is not too much of a problem. It is similar to tracking the cursor on a computer screen and clicking the mouse with your finger. Nonetheless, it is a less elegant solution, especially during flight or in combat and the hands are required for another task.

4 Comments

Great New York Times article on Social Robots

Posted by jasonwong in interruptions, robots, social on July 29th, 2007

Robots have long been a sci-fi favorite, but the robots we traditionally think about have been a failure. Sure, we have robots that can search through debris and do incredibly dangerous or tedious jobs that humans don’t want to do. This is good for humanity, but this is not the general-purpose robot we all think of. The word robot was coined in 1942 (Wikipedia), but what robots do we have in general use today? We have the Roomba… and that’s about it. The attempt to create a robot with sophisticated artificial intelligence has not been successful. We don’t understand enough about human cognition to try and implement it in an artifical being.

Well, there’s been a new push in robotics recently to develop so-called “Social Robots.” They do not take care of us – instead, we interact and take care of them. These robots are not a one-way butler or maid. Instead, we interact with them, and then we develop an affinity for them. They may not be able to go grocery shopping for us, but the can still act on our behalf.

I’ve mentioned how detrimental interruptions are to our workflow before, and researchers are working on metric to determine how engrossed we are with our work – keypresses or mouse clicks/minute, eyetracking, etc. I personally would be disconcerted if my Mac’s built-in camera was watching me to determine if I was busy. However, a little desk trinket that appears to be a cute toy could easily earn a place on my desk and also monitor my keypress rate and even my eye movements. If it were squishy, yellow and adorable (see video below), I’d be more likely to adopt such a monitoring technology. This is all about tapping into some kind of social need. Instead of trying to get robots to mimic our cognition, we use our understanding of the human psyche to build robots that we want to interact with. It’s a fascinating example of understanding how humans function and designing a system around that.

The New York Times has a great article about this new field, which you can read about here. Or, if the link goes away, you can read a PDF here.

So who in the world cares about social robots? They certainly can’t do your laundry, wash your dishes, or walk your dog. But these social robots hit some intrinsic need. Perhaps, a need…. to dance:

No Comments