Better Cheese with Machines

Who would have known something so innocuous as cheese would be breaking so many boundaries. I talked last year about the cheese wheel that went into space. Now some research in the Journal of Food Engineering offers some methods for measuring cheese composition. You know an article that begins The cheese industry demands... is going to be good. I picture them to be a 1950s-esque men in black kind of organization setting the requirements for cheese.

But food composition is serious business, after all Taco Bell is being sued for its beef content in its tacos. Using ultrasonics to detect food composition has been around and has advantages over traditional methods. Mostly you don't have to destroy the food you're testing. In this case the experimenters were analyzing eight brands of cheese. They extracted cylinders of the cheese samples and used a pair of narrow band ultrasonic transducers. These devices generate sound waves in the ultrasonic range and in this case send them at the cheese. They rounded out their test equipment with a pulser receiver and a good old oscilloscope.

They then measured the ultrasound velocity of the samples at various temperatures. The receiver measured the time it took the wave to travel through the cheese, six trials each, and they had the cylinders measured to an accuracy of 0.01 mm. What maybe was expected was that the velocity would be based on the fat and water content of the cheeses. The fat behaving differently at various temperatures was what led to velocity measurements at different temperatures.

What they found was the structure of the cheese mattered a lot. The equations for modelling wave propagation based on dairy fat could change based on how the manufacturer made their cheese and the texture of the cheese even if the fat and water composition was very similar. The experimenters also made their own cheese blends using a vacuum to remove extraneous air and found their predictions fared better in this case because the differences in cheese structure and manufacturing didn't have as much of an effect.

So sound waves to test food? Probably not too far around the corner. It's easy to think of industrial applications where this would simplify your quality process or even better a way of making government food agencies much more productive with their limited resources. But there would have to be a lot of development and specialization to get it to be accurate.

Telis-Romero, J., Váquiro, H., Bon, J., & Benedito, J. (2011). Ultrasonic assessment of fresh cheese composition Journal of Food Engineering, 103 (2), 137-146 DOI: 10.1016/j.jfoodeng.2010.10.008

A welding robot for all those hard to reach places

Many of the challenges in ship building or ship maintenance stem from environmental concerns: from the seemingly minor in not wanting ship material or ship waste to enter to the ocean to the obvious preventing future spills or catastrophic failures like the Exxon Valdez. One of the solutions to making liquid-cargo ships more resistant to failure is a double hull design.

When major disasters like mining disasters or situations like the gulf oil spill occur one question people often ask is why we still have humans doing these dangerous jobs. And that comes into play with the welding necessary in assembling a double hull ship design. The space welders have to be in to do their jobs are very small with floors and girders blocking them in and the areas can get extremely hot during the day. All this combined means the work can be extremely dangerous to the people involved. Attempts have been made before to design welding robots that could do the work in place of human workers.

Welding carriages exist but these are limited by often being able only to operate within a few axes of motion. Larger and more flexible 6-axis machines exist but the overhead devices needed to hold these robots don't fit within the confines of a double hull structure. The authors of the study found a viable commercial option that is small and lightweight and able to operate within the small area required but still unable to make the important u-shaped welds often required in these spaces due to limited control and degrees of freedom of the robot.

The designers took the leading commercial welding robot design and came up with some additions to optimize it. One issue with getting the welding robot into the space is the small size of the access hole combined with a heavy and large robot can make it near impossible for human hands to guide it in. They developed a bridge plate that supports the weight of the robot and along with two winches would allow two workers to move the robot in and out of the access hole.

The final design uses both driving wheels on the floor along with passive wheels for guidance overcoming previous design weaknesses that relied on driving wheels only and could not move as accurately with dirt and debris on the ship floor. The rails it drives along fold up to allow it to move in and out of the closed space. It contains a six axis welding manipulator, comparable to other high end commercial devices, along with a sophisticated controller and positioning device. Six AC servo motors (encoders) are directed by a four layered logic modularized controller in the CPU.

Designers also reduced the size of the hand held controller needed to operate the robot as well as adding in the capability of it communicating wirelessly with the robot rather than via a cable. In the end they reduced the time required to weld, still met satisfactory welding standards, and based on field testing have some ideas for future design.

As in many other applications, some of the best developments come out of combining previous successful designs or seemingly slight improvements to the existing solution. This is also one case where field testing proved incredibly useful and I'm sure we'll see some interesting follow up from this study.

Lee, D., Ku, N., Kim, T., Kim, J., Lee, K., & Son, Y. (2011). Development and application of an intelligent welding robot system for shipbuilding Robotics and Computer-Integrated Manufacturing, 27 (2), 377-388 DOI: 10.1016/j.rcim.2010.08.006

All about the attitude

Does the old saying fake it old you make it hold any water? Turns out maybe. Researchers from Columbia and Harvard Universities posed subjects in one of four positions: two high power positions(expansive, open limbs) and two low power positions(contractive, closed limbs). Then they measured risk taking, self-response about feelings, and testosterone and cortisol.

The high power positions were sitting stretched in a chair with legs propped up on a table and arms behind the head as well as leaning "confidently" over a table while standing. The low power positions were sitting in a chair with hands folded in the lap or standing with arms wrapped and legs crossed. I've helpfully recreated these in poorly drawn stick figures.

A saliva sample was taken 10 minutes after the subject arrived as a baseline. And then later, 17 minutes after they had held their pose. They held each pose for 1 minute and were asked to focus on faces that were shown while holding the position.

To measure risk taking they were given $2 and told they could either keep it or roll dice for a 50/50 chance of either losing it or doubling their money to $4. To get self-reported data they were asked how powerful and in charge they felt on a scale of 1 to 4.

So what happened? The stress and testosterone showed correlation in that high power position resulted in higher testosterone and lower stress and the low power was the other way around.

Of the high power posers 86% took the roll of dice versus 60% of the low power posers. The mean higher power self-rated "in charge" on a scale of 1 to 4 was 2.57 versus 1.83 for low power (both with a standard deviation of 0.81).

So clearly there's something to this. But what exactly? With 42 participants I feel like the sample size is good enough to draw some conclusions. But it'd be nice to separate some of this out. Did taking a risk make them rank themselves as feeling more powerful? Is there any reason to believe the constricting positions were just less comfortable and possibly increased the stress hormone as a direct result? Still it seems to show the possibility that body posture can affect your own level of perceived confidence and a strong likelihood that it can cause slight fluctuations in your hormone levels. So maybe you can fake it until you make it.

Carney, D., Cuddy, A., & Yap, A. (2010). Power Posing: Brief Nonverbal Displays Affect Neuroendocrine Levels and Risk Tolerance Psychological Science, 21 (10), 1363-1368 DOI: 10.1177/0956797610383437

Want to see more of my writing? Check out Engineer Blogs where I and several other engineers blog about- well, what else, engineering.

Computers and the Homeless

Everyone knows there's a big gap in worldwide between the have and the have nots in one area: internet access. A new study looked at that gap where you would expect to see it in an obvious way. They looked at computer and internet use in homeless populations in the Philadelphia area. The results are somewhat surprising.

 

Right now 58% of households have some kind of computer and 76% of these households have access to the internet. Of the homeless population they reported an average homeless time of 12 years. Of the 100 study participants, 47% reported computer use within the last 30 days. By far the main source of their computer use was free access at the public libraries in Philadelphia. University libraries, service organizations, churches and coffee shops rounded out the other sources where homeless were able to access the internet.

 

Users lauded the free public library services and discussed ways in which they manipulated the system for even more time on the computers than they were supposed to be allowed. Job searching made up the majority of what they used the internet for, followed by general internet surfing, word processing, social networking, and housing searches. They cited using the internet for "social connectedness" likely a good motivator for the social networking use. Their leisure use reflected what you might expect from a typical internet user: surfing the internet, reading the news, listening to music, shopping, books, movies and sports.

 

This might have been the first study to look at the homeless alone and their technology and computer use. What are the implications of this study? The homeless themselves suggested that cities and other agencies consider using the internet to disseminate information to homeless in their communities. I think it also emphasizes how important public libraries and computer and internet access there are to homeless populations or poor or transient populations in general. It also sets the bar for what information we want to pass on to these people and gives us perhaps another outlet at reaching them. Spreading health information and the possibility for social and community groups that could support these people having a stronger presence on the internet are other possibilities this study indicates might be a good push in the future.

 

Eyrich-Garg, K. (2011). Sheltered in cyberspace? Computer use among the unsheltered ‘street’ homeless Computers in Human Behavior, 27 (1), 296-303 DOI: 10.1016/j.chb.2010.08.007

Sneakier Submarines with Actuators

Submarines are already pretty silent and deadly. Most are either nuclear powered or have one or more diesel engines. Some use fuel cell technology to reduce the number of moving parts. Once the engines power the onboard batteries, that power spins the propeller allowing the craft to go under water for long periods at a time. German U-Boats may no longer be the enemy under water but it's still crucial that submarines are able to go silent at depth and be capable of quick and quiet escapes whenever possible. Everything that goes into the shape and power of the submarine is meant to reduce its noise.

Researchers modelled a submarine hull as a cylindrical shell with two cone shells on either end. They converted the displacement components of the cylinder to wave propagation equations and used a point force representation for the propeller as a Fourier series. The goal was to reduce the sound generated in the radial direction from propeller movement both from the propeller itself and vibration and displacement of the hull as a result of the propeller movement.

The researchers proposed an array of 60 commercial off the shelf linear actuators in a ring around the aft cone forward of the propeller.

The actuators act as a mass-spring dampening system. Their effectiveness would depend on what kind of controls were used in the feedback system to power the actuators. They examined both Active Vibration Control (AVC) and Active Structural Acoustic Control (ASAC).

The researchers found actuators could be effective at reducing radial sound generated from hull vibration. Both systems of controls they used can be seen in the above figures from their study. The actuators were especially effective at lower frequencies. That's probably a good thing as propellers typically have lower RPMs than the engines that power them. Though it's hard to see the Navy attaching a ring of off the shelf actuators to the hull of a submarine it is interesting to think what the implications might be or how actuators might reduce noise in similar applications. Or if actuators can be proven to reduce the noise what more practical mechanisms might mimic actuators successfully enough to reduce the sound profile of a submarine even further.


Caresta, M. (2011) Active control of sound radiated by a submarine in bending vibration. Journal of Sound and Vibration 330 (2011) 615–624.