Encouraging patent filings by students

I had attended the Karnataka Student Paper Awards event in Bangalore on Sept 22, 07, and had a chat with some of the prizewinners. Some of the ideas I heard about were really cool, and had significant business potential. A student himself/herself might not think of a filing a patent application even when this would be a very appropriate thing to do. So, I looked up the web to find examples of university policies for encouraging student filing of patent applications. Here are some of the relevant links:

http://www.clemson.edu/research/ottSite/ottPolicies_patent.htm

http://www12.georgetown.edu/student-affairs/Intellectual%20Property%20Policy%205-4-06.pdf

http://www.hhmi.org/resources/labmanagement/downloads/moves2_ch11.pdf

http://www.senate.psu.edu/agenda/apr25-00agn/apr25-00agn.html

A few points I noted were the following:

a) Several universities/colleges are willing to cover the costs of patenting a student’s invention, if it arises from his/her work at the university.

b) Depending upon the extent of university’s contribution to the student’s patent, the university could ask for a part of the commercial benefits arising from the patent.

Obviously the educational institution could teach Intellectual Property Rights ideas and encourage students to write disclosures of inventions. They could have faculty advise the students on improving the application. The educational institution could also market the idea to potential licensees. I believe that efforts in this direction would be of great value as long as all this is done under well-written policies and guidelines, and is properly supervised. The institution could set part a sum of money annually for this activity, and spend it wisely by choosing inventions it wishes to be involved in as a party spending money on the patenting process. In any case, teaching on IPR topics, and consultation for students on their inventions could be made available separately, protecting student interests by a suitable non-disclosure agreement arrangement.

All of us interested in encouraging student creativity should share information about good examples of these practices, and on success stories of student invention.


Srinivasan Ramani

A Robot Fielder for Cricket

A number of robots bowling cricket balls have been proposed, and many have been built. But there are not many machines for fielding, which requires computer vision or some substitute, the ability to predict the ball’s trajectory, considerable skills in catching a ball in flight and effectively throwing it back to the wicket keeper or bowler. One publication you may wish to consult is

An introductory course in mechatronics: Robo-CricketWyeth, G.Mechatronics and Machine Vision in Practice, 1997. Proceedings, Fourth Annual Conference onVolume, Issue, 23-25 Sep 1997 Page(s): 20 - 25Digital Object Identifier 10.1109/MMVIP.1997.625232

What I propose here is a free running fielding robot to play with human cricketers on a real cricket field. There must be a hundred different options in building a robot like this, but let me offer one set of options for the design for the main subsystems.

The computer vision component: Three cameras strategically located around the playground, looking at the batsman, should work with a computer to identify the ball and predict its trajectory and time and place of falling. Wireless links should carry the information to the robot. There is no reason why the robot cannot carry the computer, though that is not very necessary to demonstrate the working of the basic design.

Ball capture mechanism: This would need to catch balls in different trajectories – including a high trajectory, a low trajectory and one which involves the ball bouncing off the ground in a few places and/or rolling. A funnel like device about a square meter in cross section could be useful to catch balls coming down after flying several meters above the ground. Tilting the bucket-like funnel could catch low flying balls. Catching a ball bouncing along or rolling along the ground would require a bucket that can be tilted to come close to touching the ground. A square or rectangular mouthed funnel would be ideal for intercepting bouncing or rolling balls.

Navigation: A three wheeled vehicle with three independently driven wheels could be useful. The wheels could be suspended by individual legs, each working like the nose wheel of an aircraft, being free to swivel around the vertical support axle. I believe that this could eliminate the mechanical component of a steering mechanism. Sensing the direction along which each wheel is rolling and differential drive to the wheels could provide for electronic steering. Navigation would involve running to a suitable place to catch the ball and orienting oneself to face the incoming ball.
All this would involve the robot having to “know” its own coordinates and orientation. Some form of a position sensor and a gyrocompass-like device might be necessary.

Building a system with adequate power to provide the necessary acceleration and speed would be a demanding challenge. I estimate that a human fielder manages to do this using about 600 watts of power. A vehicle driven by a large rechargeable battery could manage with less power. How many hours of fielding such a battery would cover has to be examined. Is there room for a solar panel? I do not know.

Communication: It would be useful for the robot to use a speech synthesizer and a variety of canned utterances to radio its own comments as it chases balls. A few examples of the utterances follow:

Boy! That one is a sixer!

Well, I will get that one!

He is a goner!

Now I have to run like crazy!

John, I leave this one to you!

Throwing the ball back: This might require some form of an air gun, and a pump for recharging the pneumatic tank. The computation fo the launch angle, throwing velocity and direction of throw would be a demanding one.

Why do we need to build this gadget? I do believe that this would be a great training ground for future robotics engineers. And they would have a whale of a time during the learning phase!

I believe that a channel telecasting a cricket match would love to have a fielding robot to liven up the match! Will they pay for the R & D? That is the important question.

Srinivasan Ramani

Paperwork and Loss of Productivity

There is paperwork in all economies, and more so in some than in the others. The research issue I raise here is to what extent a nation’s GDP is reduced by avoidable paperwork.

What is avoidable paper work? In using this phrase I have in mind to the cost of time and resources involved in doing what could be done in nore efficient ways, usually by re-engineering the process or by inducting relevant technology.

The loss in productivity arises in three ways:

a) People spend time in preparing applications/documents, traveling to some office, and then in getting the paperwork done there;
b) Economic activity can get delayed by avoidable paper work, for instance the construction of commercial property gets delayed by months in getting all the paper work done; import and export of goods are slowed down by avoidable paper work.
c) Some people in some countries pay “speed money” to get things done. This often creates a motivation for staff involved to slow down the process or complicate matters for those who don’t pay speed money. This compounds the loss that occurs anyway because of unnecessarily complicated processes and by the absence of relevant technology. Of course, this is only one form of corruption related to paperwork.

I will give a couple of examples from my own experience in India.

1. It was in 1999, in Chennai, when my father passed away. I traveled from Mumbai to Chennai and spent three or four days there after the cremation, trying to get done things like arranging for access to my father’s bank account for my mother. I had to make a visit to a govt office in Mylapore, Chennai to apply for a death certificate in this connection. I stood in line for about 90 minutes, and when I got to the counter they asked me for the form issued by the hospital where my father had died recording his death. I had applied for this certificate at the hospital earlier that day and they had told me to come the following day to collect it. All I wanted at the govt. office was to get a blank form to fill in to request a death certificate. I was planning to fill it up and leave it at home for someone to submit it when the hospital form arrived. I had to leave that evening to Mumbai. The clerk at the counter told me that they don’t issue application forms until the hospital certificate is shown to them. I returned home in disgust. Later, a friend told me that the clerk would have been expecting a hundred-rupee note. Hundred rupee notes are good substitutes for documents not readily available :=)
2. This year, 2007, I bought my used company car from my company. They gave me forms signed to show that they had sold the car to me. I was supposed to get the car registered in my name. This involved a trip to an office 20 KM away where the car had been originally registered, to get a “clearance certificate”; such certificates are sometimes called no-objection certificates. Nothing moves without several no-objection certificates! Now I have submitted my application for car registration to another office near where I currently live. I have been asked to wait for a week or ten days for the paper work to be done. This has involved so far one trip to the office to get the forms and to find out what supporting documents need to be submitted, and another trip to present the application. I do not know how many more trips I will need to make to get the process completed in this IT Capital of India, Bangalore!

Let us now come to my research proposal. Such research could be carried out by a graduate student in management, by a researcher in the area of productivity, or by some one concerned with administrative reform. The proposal is that economically significant causes for the loss of productivity should be examined: time lost by citizens and by staff doing avoidable or slow work, and loss created by delaying the use of scarce resources. I suspect that in India, the loss will be somewhere in the range of 1 to 5% of the GDP. That is a loss of roughly 8 to 40 Billion dollars per year. Another way to look at this issue is to note that the GDP can be increased by 1 to 5% by taking the necessary steps to correct the problem.

I have not discussed here the question of economic losses incurred by avoidable delays in legal procedures. Undoubtedly, they are also worth studying, and quantitatively estimated to the extent possible.

My wise wife read a draft of this posting and asked me a question: are you merely suggesting more paperwork, possibly resulting in a conference paper?

No, I explained, we hope that the proposed research would lead to action!

She then gave me her no-objection certificate, enabling me to post this.

Srinivasan Ramani

Why Toys Should Use Speech Communication

Consider collaborating robots communicating with each other via Bluetooth channels or something like that. I will argue that there is something far more valuable: robots using speech for communication. I have in mind mainly robotic toys, but obviously there are other applications for the techniques I discuss.

What is special about speech? Firstly, it is a medium that enables anyone and everyone in a given area to hear what one is saying. Speech is a broadcast technique. Of course, you could have wireless broadcasts using some part or other of the radio spectrum as a substitute for speech.

What is special about speech broadcasts, as against other broadcasts? One feature is that for human observers, it is so much more fun to hear robots telling each other what to do, or sharing information among themselves.

Then there is something else: speech does not require a prior arrangement or contract, at least among human beings sharing a language. You can use speech to communicate with any stranger, as long as you two share a language. If robots share a human language widely used in the region where they are sold, robots made by one company can talk to robots made entirely independently by another company. Unexpected interactions can result with very interesting results.

Yet another effect of robots using speech communication is that they can also communicate with humans. In the case of toys this would mainly be with children. Game-like situations that could be created with these techniques could have great value in stimulating children to interact, and thereby improve their communication skills.

There are other fun things with these techniques. One of them would be to have the “language skin” learnt rather than “hard-wired” (decided and frozen at the time of manufacture of the toy). By language skin I mean the relatively superficial differences between human languages, such as differences in vocabulary, sentence patterns, pronunciation and accent. One obvious value in making “language skin” learnable is that the learning robot toy would acquire the “language skin” used by the family members of the child owning the toy, promoting the integration of the toy into the family and making it easier for the family members to communicate with it.

It would also give opportunities for the child to “teach” communication skills to the toy, challenging the child considerably to develop his/her own communication skills in the process.

Of course, talking robots are not necessarily restricted to use by children. Lonely people might find talking toys at least as interesting and comforting as pets, particularly if the external appearance and behavior are carefully designed for this purpose.

Another interesting possibility is to interface speaking toys with a cell phone that could be externally visible or hidden in the toy’s body. Then the toy could communicate with people “it knows”, in secret or in public, to share information or to request and obtain information/advice.

Using speech communication with machines for education and for fun is an area with some history. Specially designed IC chips and other hardware/software items are commercially available at different levels of capability. I list three references below:

A Texas Instrument toy consisting of a speech synthesizer and a keyboard, which was demonstrated in 1978.
http://en.wikipedia.org/wiki/Speak_&_Spell_(game)

Robota: Clever toy and educational tool Aude BillardAutonomous Systems Laboratory, School of EngineeringSciences Techniques, Swiss Federal Institute of Technology at Lausanne (EPFL), 1015, Lausanne, Switzerlandhttp://lasa.epfl.ch/publications/uploadedFiles/BillardR&A03.pdf

Where to Look: A Study of Human-Robot Engagement
Candace L. Sidner*, Cory D. Kidd**, Christopher Lee* and Neal Lesh*
Mitsubishi Electric Research Labs* and MIT Media Lab**
Cambridge, MA 02139
http://www.iuiconf.org/04pdf/2004-001-0012.pdf

The Birth of the Mind

I must share my excitement over reading the above-mentioned book by Gary Marcus, published by Basic Books (see http://www.psych.nyu.edu/gary/birth.html ). It is a tour through a number of issues: cognitive development, evolution, medicine, behavior, genetics and biochemistry. The search is for answer to important questions such as:

* How can our genes, assumed to be about 30000 in number, create the infrastructure for language, thought and related mental functions? How do they determine the nature and interaction of about 20 billion neurons? Isn’t there a severe gene shortage?
* How come the chimpanzee, sharing 98.5% of our genes, has been left so far behind?
* Are there genetically determined structures in our brain giving us our unique language-using ability, in the entire animal world?
* What is the role of nature and nurture in determining the features of our mind?

I read all this from the point of view of one concerned with AI. My focus was on:

* What can we learn from all this, for making systems better equipped to use human languages?
* Can we get any insight into what is built-in and what is learnt in the domain of language behavior?
* Should we not start with less empty learning systems, if we wish to have them learn successfully?
* Should we not recognize different levels of behavior and value intelligence that copes with the limited world of every day experience?

This is not a full-fledged review of the book. So, I won’t mention all the important issues discussed in the book. I will also not give away the secrets! Go read the book, if you want the answers.

I will only share my opinions, some of which were my own opinions strengthened by reading this book; undoubtedly, there were also some new opinions triggered by the experience. Let me list both types:

* In our search to create intelligent machines with language behavior, as models of man, we have had red herrings drawn across our trail: formal methods of reasoning, completeness and consistency.
* Completeness of knowledge and understanding, formal modes of reasoning and consistency of beliefs are not the hallmarks of human abilities.
* Many researchers in AI have ignored the lessons we have to learn from biology, like:
- Our nervous systems, like those of all other animals, are special purpose computing systems with a limited purpose – competition for survival and the propagation of our genes.
- Evolution does not have the means to suddenly bring forth into the world completely new types of organs and mechanisms. Nerve cells evolved from other cells.
- The brain develops in the fetus in ways similar to that of other organs.
- Our sense of wonder only increases, as we understand in scientific terms more and more about our bodies including our brains, and our behavior.

Let me conclude by saying that I will not accept an AI researcher as fully educated if he has not read books like this!


Srinivasan Ramani

Numenta

Jeff Hawkins is the kind of dreamer who takes another shot at AI-like stuff, putting his money into it. But given his impressive record as the man behind the PalmPilot and the Treo, and as the inventor of Graffiti, I was willing to read an article about the new company he founded in 2005. See

http://money.cnn.com/magazines/business2/business2_archive/2007/02/01/8398989/index.htm

Donna Dubinsky is the CEO.

The claim is that “algorithms that Numenta has come up with allow machines to learn from observation, just as a child learns by observing the world around her”. But that is accompanied by Hawkin’s statement that “Numenta has nothing to do with the field known as artificial intelligence”. Someone would have to explain that to me! Why isn’t that AI? Possibly it borrows little from techniques known to scholars of AI. However, once you claim to build an algorithm that enables machines to learn from observation, you cannot disown AI. Hawkins claims that his work is inspired by how the brain works.

Dileep George, a Stanford Alumnus, is credited with contributions to algorithm development. An early demo, that the algorithm was used to give, dealt with recognition of animated images like those of a cat, dog and so on. The basic concept is called “hierarchical temporal memory”, involving a hierarchy of nodes for storing patterns and sequences of patterns.

The company plans to release a Beta version of the software and some development tools soon.

Is the idea going to turn out into a simple one? Even if it does, it could be a useful step forward. After all, notions like Finite State Automata, Perceptrons, and Pattern Matching Algorithms have served AI well. I for one would be waiting to read more about Numenta.

Srinivasan Ramani

Student Project Contests

This news story is about an amazing set of competition ideas for those who love to design and construct machines and/or program them. National Institute of Technology, Suratkal in India, will be hosting an event named Engineer 2007 during March 7-11, 2007. Pity it is so soon. But don't despair; read on.

Visit the site

http://www.engineer.org.in/events.php

and read about the challenges set for contestents in Robotics, Computer Science events including one on applications for a three-degrees-of-feedom mouse named Mushaca see

http://ssl.serc.iisc.ernet.in/~manohar/Research/MushacaManif/mushaca.htm

and a "Kode Kombat". There is also an Electronics challenge involving the extraction of information from a video stream.

The ideas are thought provoking; and, who knows, you might decide to take up one of these challenges just for the fun of it.

Srinivasan Ramani

IEEE Student Paper Contest 2007

The deadline for an IEEE Student Paper Contest is March 15, 2007. You can find time to write a paper even at this stage. This has been announced as a Region-10 activity (Asia Pacific), but you can check with the organizers if they accept student papers from all over the world. They probably do that, but I am not sure. Visit

http://www.ieee.nus.edu.sg/index_files/Page1386.htm

I have been looking for information on the IEEE Student Design Contest 2007. You will get a feel for this by visiting

http://www.edn.com/article/CA605949.html?industryid=2814 describing entries in Contest 2005.

I have not been successful as yet in visiting the Student Design Contest 2007 website. If any of you know an accessible website for this do let us all know.

I have heard that the Student Design Contest is worldwide, and that it provides hardware and software to do a project of your choice on a theme announced by IEEE for the year. They do not expect you to spend any significant amount of money yourself on buying components, hardware or software. All this makes the Contest a very attractive one.

Srinivasan Ramani

Research Excellence Award for Students

You should read the posting at

http://omegageek.net/rickscafe/?p=933

A nice list of criteria which are used to judge student's submissions on their research is given there. This list will be helpful while writing any research paper for publication.

Ramani

Sharing Information on Graduate Student Research Projects

I don’t think that there are many sites that encourage graduate students interested in research projects in information technology or computer science to communicate among themselves. If you know of any, please tell me. I have occasionally done a search for such sites and have not found any.

On a related note, it is good for graduate students to know about dissertations that have been written in their own areas and to get copies of them. The worldwide trend of universities maintaining an open archive to share their dissertation collections provides one opportunity. A number of US universities do that. The Australian have a national arrangement to pool information on the dissertation collections from as many as 20 universities. You might wish to look at this site: http://adt.caul.edu.au/

I found a number of theses that are very interesting, on the basis of my own interests (there are many more in CS and IT). I could read them from the document file, except in one case – the first one. Some provide a copy of the document in one pdf file, and others have different files for different chapters. The common search facilities provided by the above-mentioned website will locate information on any of the following theses that I found noteworthy:

By Jijoong Kim, Automatic Aircraft Recognition and Identification, 2005, University of Wollongong

By Joe Cronin, Design, construction and control of an industrial scale biped robot, 2005, University of New South Wales.
(This is a 500 Kg monster! From the pages I have read so far, it looks like it has not been built. But it has been designed and the designs have been tested by simulation)

By Penny Baillie, The Synthesis Of Emotions In Artificial Intelligences, 2002, The University Of Southern Queensland

By Peter Krebs, 2002, Turing Machines, Computers and Artificial Intelligence, University of New South Wales (This thesis argues that an artificial intelligence can transcend the limitations of a formal system because it can interact with the real world)

By Keven Weber, Visually guided Autonomous Robot Navigation: An insect based Approach, 1998, Curtin University of Technology (This thesis argues that a parsimonious, but robust guidance mechanism can be developed to handle two problems: corridor guidance and visual homing).

Coming back to my initial question: do we have a wiki like mechanism for graduate students to announce what problems they are working on? If not, why not? If the student does not wish to announce his approach, that is all right. He/she can merely announce the title of the proposed thesis. Students looking at a common area could at least share information, news about tools, relevant publications etc., and send each other preprints of their publications.

Srinivasan Ramani

Semantics, AI and Dyslexia

There has been a lot of theorizing about how semantic information is used to recognize the structure and meaning of sentences. I got a very interesting view of an approach to this problem by a scholar who had started out as a medical doctor, Parag C. Prasad.
He did a Ph D at the School of Biosciences and Bioengineering, Indian Institute of Technology, Bombay, completing it last year (2006). Professor Arunkumar was the thesis advisor.

The thesis offers an interesting hypothesis as to how the human mind uses a sequence of words read (call this the immediate context) to predict the word it expects to read next. Prasad has developed a simulation model in the form of a neural net and has carried out experiments on it. The attraction of the hypothesis is its simplicity. It ignores the structure of the “immediate context”, and uses the set of words in it to represent relevant information. In other words, the hypothesis is that a small bag of words you have read helps you predict the next word you expect to encounter. The bag is the Short Term Memory (STM). The hypothesis is that the Long Term Memory (LTM) associates different bags with the individual words that they predict. For example, the bag (I, like, eat) could be associated with “mango” in your LTM.

Prasad discusses results from the medical field that show the extent of degradation of reading ability arising from disease conditions such as lesions, and shows how his neural net model successfully shows similar behavior.

Undoubtedly, there are possibilities of extending the hypothesis to account for a variety of phenomena. For instance, if you have read that John likes to eat mango, you might find it easy to recognize “banana” if it appears with the same immediate context as mango. How does this happen?

I can also think of a number of research projects that depend on the hypothesis reported to improve machine behavior. Can you think of one? Do you wish to list it here? Or do you wish to carry out such a project? Post a comment here.

Ramani

Artifical Intelligence and Society

Artificial Intelligence and Society
Symposium on AI in Industry (SAII) http://sigai.cdacmumbai.in/, held on January 10, 2007, collocated with IJCAI 2007 http://www.ijcai-07.org/

I contributed to a panel on Artificial Intelligence and Society at the above-mentioned symposium held on January 10, 07 in Hyderabad. Our task had been cut out for us as Prof Raj Reddy had given a keynote at IJCAI 07, focusing on what AI and Robotics can do for society. He had also mentioned an Indian initiative to set up a center to carry out research in the field.

I hope that report on Prof Reddy’s keynote and on the discussions in the Panel would be reported on the net soon, perhaps on the websites mentioned above, or even on this blog.

The large audience we had for the panel very pleasantly surprised me. It was heart-warming to see the enthusiasm displayed in comments from the floor. It is clear that a lot of people feel that is it is worth using challenges drawn form the point of view of benefits to society as drivers to develop AI.

Let me add my own comments here. I believe that AI has a tremendous potential to contribute to education. The world has a few billion people who have not had the opportunity to get a good education. See http://www.ei-india.com/full-report.pdf

Nearly a billion people were illiterate according a UNESCO report of 2003.
http://portal.unesco.org/en/ev.php-URL_ID=10513&URL_DO=DO_TOPIC&URL_SECTION=201.html
Their fate would no longer matter after fifty years, when they would mostly be dead. But, as humanity, is this what we want to see – a few billion people live through their lives in the 21st century with a bad education? Unfortunately, mortality among illiterates has been a major contributor to the so-called “increase in literacy” in many parts of the world. I don’t want to call it increase in literacy.

I believe that the best of available technology, and all technology we can develop, would be needed to solve this problem within the available window of a couple of decades. Of course, mere technology will not do. We will need science, particularly cognitive science and educational psychology. Even more, we will need the political will and the resources that would be required to mount this “grand challenge”.

AI’s potential in assisted living and in helping the sensory handicapped have been pointed out by many people. I will not dwell on them.

Let me give a couple of practical examples to justify my faith in the potential of AI. Dr William McBride, an obstetrician, had published a letter in The Lancet, in December 1961, pointing out that the drug Thalidomide given to pregnant women could cause serious birth defects in children. I have heard an Australian scintist say in a talk that McBride had used a very simple form of cluster recognition in his detective work to track down the cause of birth defects in children he had encountered in unexpected numbers in Sydney. In a great example of the power of epidemiology, he had used a map of Sydney and pinned a small paper flag at every address at which he knew of a child being born with birth defects. Unfortunately, soon enough, he had noticed two big clusters forming on the map. He is said to have tracked these down to two doctors who had prescribed the then relatively new drug Thalidomide to treat nausea in pregnancy. McBride won a Nobel Prize for his discovery.

(I cannot find supporting evidence for the cluster reognition part of this story, and would be grateful to readers for commenting on the accuracy or otherwise of this part).

The other example deals with a case in which simple techniques of analysis were helpless.

I quote Chris Darnell, who is now involed in very large scale financial management:

“I had been discouraged with the failure of traditional statistical methods to analyze financial market data characterized by a low signal-to-noise ratio. After reading some of the early papers by Professor Kanal in the new field of statistical pattern recognition, I called Laveen--out of the blue … He gave me some great suggestions for using new analysis methods to analyze financial data”.

Among other things, Darnell and Kanal explored the use of the open source statistical pattern recognition software ISPAHAN. This was all circa 1981, when people thought that pattern recognition was pretty much AI! Darnell has now acknowledged Prof Kanal’s guidance with a handsome endowment. The University of Maryland announced recently the Darnell/Kanal Professorship in computer science. See

http://www.cs.umd.edu/newsletters/InsideCS_2006f.pdf

I hope that these two examples provide some support to the faith that AI can do some good to the world. “Failure of traditional statistical methods to analyze financial market data characterized by a low signal-to-noise ratio” was a big problem in financial management. Can you imagine a similar problem in epidemiology? Do not many AI researchers have tools to help in this? This may not be a new idea to those working on machine learning. See

http://dimacs.rutgers.edu/Workshops/DataMiningTutorial/dmtutorialreport.pdf

However, that does not detract from the argument in favor of AI research being aimed at social good.

P. S. You might wish to read

http://newstudentresearch.blogspot.com/2006/08/exploiting-search-speech-recognition.html

for some suggestions on creating socially relevant systems using speech interfaces.


Srinivasan Ramani