The president of the small to medium-sized machine builder had always made himself a reasonable living from developing custom-based vision systems to inspect one particular type of widget.

And although the market for such widget inspection systems wasn't all that large, his system, or variants of it, had been purchased and widely used by most of the widget makers in the industry.

Recognizing the fact that the Internet might provide his company with more exposure, the President decided to hire a developer to create a web site that would explain to any new potential customers the capabilities of his company.

And that's exactly what he did. Prior to developing the web site, however, a member of the web site development team went to visit the outfit to find out more about the system. Once there, he was treated to an hour long dissertation by the company's marketing manager who explained to him exactly why there was a need to inspect such widgets and a very brief description of the system that they had developed.

After the meeting, the web developer went back to his office and created a stunning web site for the company. Not only did the web site provide a background of the company and its university origins, it also detailed the inspection problems faced by the widget manufacturers. Unfortunately, however, there was only the briefest description of the system itself, the functions it performed and its performance – all listed in bullet points alongside a rather sorry-looking photograph.

When the web site was launched, the President was confident that it would offer the world an insight into the capabilities of his company and lead to a number of new leads, not just from companies involved in widget manufacturing, but other outfits that might be faced with similar inspection problems.

Sadly, of course, that didn't happen. Having won orders with most of the widget makers already, the website attracted no new customers whatsoever. The president and the marketing manger were disappointed – not in the least because they had spent a considerable amount of money developing it. And they were both at a loss to understand why the reaction had been so poor.

Some months later, a journalist from a magazine that covered the field of vision systems design came to call upon the company. Unlike the previous interview, however, the journalist grilled the president to discover exactly how the system had been designed. He went away confident that his description of the hardware chosen for the system and the software that had been written for it by the engineering team would prove a hit with his readers.

And it was. When the article was published, it became immediately apparent to many of the engineering readers how the company could tweak the widget-inspection system to help them inspect their own products too.

The president is now a happy camper. Having received several enquiries from some potential new customers, he is now looking forward to expanding his business into new markets. More importantly, however, he has at last recognized the importance of publicizing the technical capabilities of his technical team rather than just promoting a single product line.

A team of researchers from the Stevens Institute of Technology (Hoboken NJ, USA) have developed a vision-based system that will enable students to find a seat in the university’s library during the busiest periods of the semester.

“We noticed that we and many of our fellow students spent precious study time looking for seating in the S.C. Williams library,” says team leader and computer engineer Richard Sanchez.

But rather than grumble and head back to their dormitories, computer engineer Sanchez’ team decided to develop a system to solve the problem once and for all. Called Seatfinder, it's an innovative way of detecting what seats are available in the library that uses image processing to identify the presence of an individual.

With the assistance of Professor Bruce McNair, Distinguished Service Professor of Electrical & Computer Engineering and the Stevens IT department, the team deployed an IP camera with network connectivity to capture a live feed from the library.

After capturing images of the seating, the IP camera transmits a live video feed over a wired network to a remote computer using an open source application called iSpy. Next, a motion detection algorithm is used to trigger a snapshot of the live camera feed any time an individual leaves or enters a table in the library.

The team's code then processes the captured image to determine which areas are free or occupied, after which it updates a website with an image that represents occupied and empty chairs around the table. By checking the web site, other students can then find a space to sit and study a lot more quickly.

Having successfully proven their system, the Seatfinder team now hopes to add more cameras and monitor more tables to increase the sophistication of the system. Eventually they envision deploying it at other venues where space can become scarce, such as restaurants, movie theatres or parking lots.



But I can see even more opportunities for the system, one of which is on public transportation. If you have ever traveled on any form of rail transportation in any large metropolitan area during the rush hour period, for example, you will know what a problem it is to find anywhere to sit.

While the Seatfinder system can't obviously produce more seating on a full train, if some of our rail companies could deploy such a system in each of their carriages, it would reduce the time that hapless commuters spend walking up and down the carriages in the desperate hope of finding a spare seat. Those commuters lucky enough to be able to access a website from the train, that is.

Many years ago, whilst working on an electronics journal in the United Kingdom, I was invited to a press conference to witness the unveiling of what was claimed to be the next big thing in electronic systems.

To ensure that as many folks turned up as possible, the organizers of the event decided to hold the conference at one of England's famous race tracks,  where they invited the press to remain after the company presentations to enjoy the rest of the day betting on the horses.

As it transpired, the press conference itself turned out to be a crashing bore -- the system itself had already been launched months earlier in the US, and most of the press knew about it already. Sadly, the view of the racetrack wasn't much better. You see, although the organizers had erected a tent as close to the racecourse as possible, the view from it was somewhat restricted. The result was that the horses could only been seen for mere seconds as they raced by.

Now, I'm pleased to say, a solution to this problem is a hand, thanks to a couple of savvy students from the University of Arizona (Tucson, Arizona, USA) who have come up with a solution based, of course, on the use of a vision system. That’s right. David Matt and Kenleigh Hobby’s new 'jockey cam' is a smart camera-based helmet that can stream real-time video from a jockey's head, putting the viewers right in the saddle rather than stuck by the side of the track.


The two entrepreneurs have even launched a new company called EquiSight to market the system, and have captured the attention of ESPN, the Ireland Tourism Board, racetracks around the world and venture capital investors.

Since starting the company, it's been a whirlwind ride for the pair. In December 2011 they presented their system to more than 600 racing and gaming executives at the 38th Annual Symposium on Racing and Gaming in Tucson. In February 2012, they filmed 30 jockey-cam videos at prestigious race tracks and training centers on the East Coast. And in March 2012, Wasabi Ventures (San Mateo, CA, USA) selected EquiSight to receive venture capital support.

EquiSight, which now holds three provisional patents on its technology, also recently inked an agreement with an engineering design firm to explore the potential application of helmet-cam technology for the military and law enforcement.

As a member of the press, of course, I'm now looking forward to the day when I'm invited to take a look at the system at a press conference here in the good old US of A. Needless to say, if it's held at a race track, it's bound to be more enjoyable than the last press conference I attended at such an event all those years ago.

When I was a little lad, there was nothing I enjoyed more than playing in a sandbox in the back yard of my parents' house during the summer. That old sandbox became a place where I could construct my own virtual worlds filled with castles, mountains, river and oceans. It was, as I recall, jolly good fun.

But in this age when computers dominate most every aspect of our lives, it should come as no surprise that even the humble sandbox has now been transformed into a digital experience.

That’s right. As part of an NSF-funded project on freshwater lake and watershed science, the good folks at UC Davis (Davis, CA, USA) have created a sandbox that allows users to create topographic models by shaping real sand which is augmented in real time by an elevation color map, topographic contour lines, and simulated water!

While it sounds like great fun for kids of all ages, the sandbox hardware built by project specialist Peter Gold of the UC Davis Department of Geology has actually been developed to teach geographic, geologic, and hydrologic concepts such as how to read a topography map, the meaning of contour lines, watersheds, catchment areas, and levees.


To do just that, the system makes use of the Microsoft Kinect camera which continuously collects images from the sandbox as the user interacts with the sand. The 30 FPS images from the camera are fed into a statistical evaluation filter which filters out moving objects such as hands or tools, reducing the noise inherent in the Kinect's depth data stream, and filling in missing data.

After some software processing, the resulting topographic surface is then rendered by the projector suspended above the sandbox, with the effect that the projected topography exactly matches the topography of the real sand. The software uses a combination of several OpenGL shaders to color the surface by elevation using customizable color maps and to add the real-time topographic contour lines.

At the same time, a water flow simulation is run in the background using another set of GLSL shaders. The simulation is run such that the water flows exactly at real speed assuming a 1:100 scale factor, unless turbulence in the flow forces too many integration steps for the driving graphics card to handle.

The researcher say that the software they developed to create the virtual sandbox is based on the Vrui VR development toolkit and the Kinect 3D video processing framework. For those wishing to develop their own sandbox, they say that the software will soon be available for download under the GNU General Public License.

That's surely good news for all those of us who would love to take a trip back to our childhood by sampling the new virtual delights of the sandbox in our own living rooms.

More information on the augmented reality sandbox can be found here.

There's no doubt that autonomous robots fitted out with vision systems can perform some pretty useful tasks. For the military, such robots can help prevent injuries in the battlefield by providing soldiers with a remote insight into the nefarious misdoings of the enemy.  On civvy-street, they can be used for the equally important purposes of improving the environment or keeping the borders of countries safe.

But these conventional robots are predominantly made of rigid resilient materials, many of which are non-biodegradable and have a negative impact on the natural ecology.

That means that any robot deployed in the environment must be continually tracked and, once it has reached the end of its useable life, must be recovered, dismantled, and made safe. But there is also the risk that the robot will be irrecoverable with consequent damage to the eco-system.

Now one might think that there's not a lot that can be done about this. After all, the computers, power sources and imagers used in such robotic devices are all man-made, and many of them are composed of some pretty toxic substances. And there doesn’t appear to be any alternative to using them.

But apparently, the academic folks at the University of Bristol (Bristol, UK) think differently. They believe that it might be possible to build robots that decompose once they have reached the end of their mission. While it all might sound a bit far-fetched, the idea has won Dr. Jonathan Rossiter, Senior Lecturer in the University of Bristol’s Department of Engineering Mathematics, a two-year grant of over £200,000 from the Leverhulme Trust (London, UK) to work on developing robots that rot.



That's right. Dr. Rossiter, together with Dr. Ioannis Ieropoulos, Senior Research Fellow at the Department of Engineering, Design and Mathematics at the University of the West of England (UWE, Bristol, UK), aim to show that autonomous soft robotic artificial organisms can exhibit an important characteristic of biological organisms -- graceful decomposition after death.

Since there will no longer be the need to track and then recover such robots, the deployment of large numbers of biodegradable robots in the environment will become inherently safe. Hundreds or thousands of robots could therefore potentially be deployed, safe in the knowledge that there will be no environmental impact.

Now I'm sure that there are some among you who believe that Dr. Rossiter plans are all a bit pie in the sky. But I'm not one of them. This year, for example, we have already seen the development of micro lens arrays produced by a mineral precipitation by researchers at the Max Planck Institute of Colloids and Interfaces (Potsdam, Germany).

Nevertheless, I'm going to be very interested to see if and how those UK researchers can build an entire robot that is totally biodegradable. I guess we'll just have to wait a couple more years to find out.

For more information on the work of the researchers at the Bristol Robotics Laboratory -- including a robot powered on a diet of flies -- click here.