Thursday, November 13, 2014

Landing a probe on a comet (67P/Churyumov-Gerasimenko)


The European Space Agency's Rosetta Spacecraft has succeeded in landing a device carrying instruments on the comet
The extent of success of the mission is not clear as I write this piece on Nov 13, 2014 as seen from this report:

According to TV news, gravitational acceleration at the place of landing is 100,000 times weaker than on earth; if this is true, it would mean that a 100 Kg lander would behave as if it weighed 1 gram! It is obviously a great challenge to a build a robotic device to work under such circumstances, land on a surface whose properties are not known, and then carry out useful research. All this has to be done with 64 hours of battery life! We do not know how well the solar panels would be able to recharge the batteries!  

The goal is undoubtedly a worthy one. We should recall the great contribution of Edmond Halley in this context. He predicted that a comet that had appeared in 1456, 1531, 1607, and 1682 would return in 1758. He did not live to see this happen, but the prediction made a big impact on human thought. Superstitious beliefs that comets were heavenly messengers foreboding pestilence, famine and death were shown to be hollow. His prediction was not a casual one. He had been attracted to Kepler's laws of planetary motion and was looking for proof that they were in fact accurate. In August 1684, he went to Cambridge to discuss this with Sir Isaac Newton, and found that Newton had solved the problem, working on the orbit of comet Kirch. Finding that Newton could not locate a record of his own work, Halley encouraged him to redo it and publish it. It finally came out prominently in “PhilosophiƦ Naturalis Principia Mathematica authored by Newton and published at Halley's expense in 1687.
Halley and Newton did not merely change our knowledge of mechanics. They helped destroy the cobwebs of the human mind, giving us an opportunity to develop a more verifiable world view. We don’t need to cook up myths to explain what we do not understand. We need to explore and find verifiable truths to share.

The European comet lander Philae is humanity’s latest step in this direction. As a robot it looks tragically ill equipped to search for the presence of amino acids on a 4 km rock floating hundreds of million miles away. It will no doubt inspire better robots to be built to explore heavenly bodies. The presence of certain amino acids on comets would tell us more about the origin of life. Any discoveries in this direction would take us farther along the revolution Halley and Newton had launched.  

Let us recognize the engineering challenges in this, and encourage students to design good robot explorers. It is not a job for a single team. Hundreds of ideas need to be explored and applied to solving hundreds of problems, apart from building comet explorers. I had written a piece on a related topic in 2012 and the URL for that is    

Srinivasan Ramani




Tuesday, November 04, 2014

A big asteroid discovered – 2014 UR116


An automatic telescope in Russia near the city of Kislovodsk has spotted an asteroid not sighted earlier. It has been now named 2014 UR116. This asteroid, estimated to be 370 meters in diameter, does not pose any threat of possible collision with the earth over the next six years or so. However, its orbit gets disturbed by the influences of venus and mars. So, there are no guaranties for the future! Obviously, earth's gravitation also plays a part in altering the object's orbit. Visit


Would it not be great if India builds an automatic telescope which makes available shots of specific areas of the sky on requests over the Internet? Many of us might spend some time looking for interesting astronomical phenomena if such images were easily available.

Srinivasan Ramani



Saturday, September 13, 2014

A product idea for a startup – A Traffic Sign Recognizer

Product ideas that are waiting for a scientific advance are not ideal for most startups! You cannot get such advances merely by putting pressure on the staff! Nor are ones requiring a whole lot of technical R & D. In contrast, there are ideas tried and tested as reported by a set of papers.  One such is a traffic sign detector and recognizer. Can you visualize a car mounted video camera that looks out for traffic signs such as stop signs, speed limit displays, “don’t enter” signs, children crossing warnings? A number of researchers have shown this can be done. Visit
Road and Traffic Sign Detection and Recognition
http://www.iasi.cnr.it/ewgt/16conference/ID31.pdf
One announcement on the Web claims that the reported solution performs at superhuman level.

Use of colour for detecting and segmenting the traffic sign image from the overall scene is quite popular. Use of Support Vector Machines is a popular technique and so is the use of neural nets. It is clearly within reach to detect and recognize traffic signs from good quality photographs, and to give lab level demonstrations of the techniques.

Use of visual pattern recognition offers the big advantage that a product does not have to wait for the world to be re-organized. Traffic signs that are in wide use for human recognition serve as the environment for camera based techniques. Imagine asking a city to add even one LED to each traffic sign and to ensure regular power supply to it! It would be a near hopeless task. Camera based techniques do not require such re-engineering of the environment.

A question that arises is why is there no product of this type in the market? Is it because of legal worries – a guy land up in an accident and file a suit saying that a product weakness caused the accident. I don’t think that this is a major issue. Proper labeling and use of Terms and Conditions approved by a lawyer would safeguard the tool. Early versions of the product can be identified as being purely advisory and the user can be cautioned not to depend upon their accuracy and completeness in recognizing signs.  

An important research goal is to use two or more independent recognition techniques in one device and see how the results can be used together to improve product accuracy and reliability. Work on a prototype should be treated as a software engineering project, in which product reliability and performance are given importance and bugs are meticulously tracked and fixed. Optimizing the code to achieve quick response is obviously important too, but nothing can substitute for a well-thought out clean design to start with.

A camera based recognizer can be built as an app for a cell phone in the first place. This will also give it access to GPS coordinates of the vehicle it is mounted on and Web-based access to a remote server. It is one thing to recognize a stop sign at an intersection; it is far better to anticipate that traffic signal before the car comes within sight of that intersection. Integration of a map on the server with the traffic sign detector and recognizer will advance the state of the art to a superior level as compared to that of an unconnected sign recognizer.

Srinivasan Ramani       

Thursday, June 19, 2014

Getting an auto rickshaw at your doorstep!

This blog has discussed this issue earlier and has suggested a few related project ideas: 
The purpose of this posting today is to suggest a simple project you can implement in a day or two, excluding a few days of learning time if that is necessary in your case. I will describe it in the Indian context mentioning “autos”, that is auto rickshaws. Obviously it will work for a variety of vehicles such as cabs, and even the inter-city buses that wait for passengers!

The suggestion is this: Create a website at which auto drivers can register, giving their name, cell phone number and vehicle number. Later, they should be able to send a short message to a server, giving the location at which they are waiting for passengers. They should mention a clear landmark, for example: Lal bagh.

The site will display over the Web a table giving driver’s phone number and location giving city name as well. Any one requiring an auto could do a web search mentioning something like this:
Site:www.getauto.in   Lal bagh
and select one of the displayed phone numbers. The rest of the work of contacting the driver and giving him the pickup address is left to the user. Later on one can add additional facilities like sending the user an SMS/email giving driver name and site registration number (SRN). In the rare event it is required, the police can use the SRN, contact the site and get driver’s license number, etc.

Remember – only the simplest ideas work first! You can of course write an App, but wait.Should you expect the auto driver to have 2G or 3G connectivity? Or should you try to build a system with an SMS interface? I suggest that you first focus on using a 2G/3G connection. Postpone your market expansion program for the present.

Srinivasan Ramani

Sunday, June 15, 2014

Putting a radio beacon on an asteroid


The asteroid named "the Beast" is approximately a thousand feet wide, and is big enough to release 2,000 Megatons of energy if it collides with the earth. That could wipe out a whole metropolitan city. Fortunately, it did not do that during its latest pass on June 8, 2014. It is not expected to come dangerously close any time soon. Officially, its name is 2014 HQ124. For further information, here are a couple of URLs. 

http://www.jpl.nasa.gov/news/news.php?release=2014-178

This time, the Beast passed within 750,000 miles of the earth. Isn’t time that a technology is developed to attach a transponder to such asteroids so that they can be tracked accurately as far away as a hundred million miles? A good solution would be to have the transponder orbit the asteroid in a fairly tight orbit. It would be necessary for the transponder to orient itself in space and point an antenna towards the earth for long distance communication. Would the development of such an orbiting transponder be a good project for university students?

Why a transponder? Why not a simple beacon? A transponder allows fairly accurate distance measurement; but that may not be essential. A beacon that transmits at fixed intervals may be sufficient to let good tracking of the orbit to be
 done. I think that a major challenge is in designing a system to locate the direction of the earth and to keep a directional antenna pointed at it. Are there powerful signals sent out 24/7 from the earth that can be utilized to sense the direction of the earth from the asteroid? The Wikipedia gives a list of radio time signal stations around the world, including a 2000 KW transmitter working at 162 KHz in France and a 2.5 KW transmitter working at 20 MHz in the US. Obviously, the 20 MHz transmitter is attractive because one can design a highly directional antenna to work with it, but how do you design a simple mechanism to locate it in the first place and to re-acquire the signal if it is lost for a short time for some reason? One option seems to be to use a Sun seeker and to use rotation with an axis perpendicular to the plane of the solar orbit to acquire the earth signal. Assuming that the asteroid's orbit and that of the earth are not highly inclined to each other, the directional antenna sweeping the plane of the orbit can locate the earth. The satellite can then de-spin itself and point the antenna to wards the earth. The system should be able to work with a few different radio sources spread over the world.  


Another challenge would be in increasing the design life of the transponder to make it worth the effort of launching it. Yet another challenge would be in having the transponder adjust its orbit over the years, so that it does not crash, or get pulled away by the earth or the moon during its future passes. Both challenges would require that one does not depend upon rocket fuel being carried on board, because it would run out rather soon. Considering the very low gravitational force involved, ionic propulsion is a possibility for making necessary changes to the orbit. Or, who knows, may be a good solution is to launch a one ton tank of fuel, with a 20 Kg transponder tacked on outside!


Srinivasan Ramani

Saturday, May 03, 2014

John McCarthy, Guru of Artificial Intelligence - Read about him in Resonance


Everyone in computer science knows a lot about John McCarthy and his contributions.  LISP was the first programming language I fell in love with. His contributions to AI and computer science are well-known, but I confess that I have never read his biography so far, but I caught up a bit recently when Professor V Rajaraman wrote a very interesting article about John McCarthy in the March 2014 issue of Resonance, which is available online. Visit
I downloaded the PDF version of the article and read it. As a bonus, there is another article in that issue by Dr Raman Chandrasekar on the famous AI system named “Jeopardy!” which competes in TV quiz programs:  http://en.wikipedia.org/wiki/Jeopardy!

Let me offer a few tidbits to give you a feel for Rajaraman’s article. He mentions McCarthy’s visit to IIT Kanpur and says it inspired a number of students to take up computer science and that one of them was Narayana Murthy! He also mentions that the parents of McCarthy had been immigrants to the US, one from Ireland and one from Lithuania. Both had been members of the American Communist Party.

McCarthy had done seminal work in Artificial Intelligence, brought researchers together in the beginning of work in the field; no wonder Rajaraman calls him Father of Artificial Intelligence. McCarthy spent three years serving as Assistant Professor of Mathematics at the Stanford University, but was not given tenure! I got reminded of young colleagues in India who are very interested in inter-disciplinary research of potentially great significance, but are scared that promotion committees may not do them justice if they “stray” beyond the mainstream!  

I do not wish to say any more about McCarthy’s ideas and work in this blog post. I only wanted to attract a few more readers to enjoy Prof Rajaraman article and to benefit from it.  He has done a great job of elucidating the important ideas that McCarthy had worked on.


Srinivasan Ramani 

Wednesday, April 30, 2014

Building a WiFi “Satellite”


This post is really a companion to my recent post 
http://www.obvioustruths.blogspot.in/2014/04/use-unlicensed-wifi-spectrum-in-rural.html Please read that if possible, before reading the following. It will give you the context in which I write. 


Here, I really mean by “satellite” a device or a system that is remotely located, and performs a communications relay role. It should be designed to withstand pretty harsh environmental conditions and run unattended. It should be self-sustaining in terms of power, depending on solar panels. It does not have to be in earth orbit! It would be sufficient if it is on some tall peak like Doddabetta in Ooty, Mullaiahnagiri in Chikmagalur, Anamudi near Munnar, Perumalmalai peak in Kodaikanal (just to name a few peaks in South India). If there is a tall building visible for miles around, it may even be on top of that building. The harsh environmental conditions that have to be withstood would be rain, lightning, wind force, heat and cold. 

Actually the system does not have to be on the tallest peak or building in the area, but on a well-chosen place that provides effective line of sight with suitable user groups like schools/colleges. The altitude at which it is located should help it to command a large area around. I expect that it would be easy to find many places in India, where one such device would cover a thousand square kilometers within WiFi range! The “satellite” should, of course, be mounted on a tower or something like that to prevent theft!

The system would include a router, several 5.8 GHz radios and directional antennas. It would have maintenance-free storage batteries and solar panels. Designing such a device is an inter-disciplinary engineering activity and I hope students from different disciplines would volunteer to look into the different aspects. Many design issues arise – like should we decide in advance what the user groups should be and fix the directional antennas appropriately? What is the ideal trade-off between simplicity and sophistication? I will leave to readers to find such issues and discuss them.

Srinivasan Ramani 
April 30, 2014

Friday, April 18, 2014

Prof R Narasimhan's Birthday (April 16)

It is Professor Narasimhan’s birthday today (April 16). If he had lived on, he would have been eighty eight today. There is one simple way many of us can further a goal he worked for - running high quality conferences in the computer field in India. His family has made available to the Tata Institute of Fundamental Research, the Institute where he spent his working life, a handsome endowment for an annual memorial award to bring a world-class invited speaker to international conference in the computer field in India. We should spread this information among program committee members of such conferences. For details, please visit
http://www.tifr.res.in/~endowment/index.php/endowment-awards/prof-r-narasimhan-lecture-award

I will quote a few lines from this site:

2. One award will be given annually in the field Computer Science and Technology and will recognize advances in hardware, software, theoretical aspects of computing, or applications of computing.
3. The award will recognize the achievement of younger professionals only and hence will be restricted to professionals up to the age of 40.
4. Recipients of the award should hold a degree from an Indian University, excluding honorary degrees.
5. The award will be in the form of a scroll, with a trophy or plaque (rather than in monetary form), and an invitation to give a public talk titled the "Professor R. Narasimhan Memorial Lecture" on the topic for which the recipient is being recognized. In cases wherein the recipient cannot attend the conference in person, arrangements will be made for the recipient to present the lecture through video conference.
6. The award will cover the expenses of the recipient to travel to the chosen conference in India and present the talk.
I wish to share a few thoughts here:

Obviously, only established international conferences with high standards will qualify for these awards. Peer reviewing and published proceedings are essential.

Prof R K Shyamasundar, TIFR, or I can offer any clarifications required.

Srinivasan Ramani

Saturday, March 22, 2014

The need to invent Secondary Data Recorders for Aircraft

The tragic incident of the Malaysian Airlines MH370 has triggered a worldwide question. Why don't we have better data and cockpit voice recorders? Rep. David Price, D-N.C., is reported to have said that the 9/11 Commission recommended after the terrorist hijackings in 2001 that planes carry ejectable "black boxes" to make them easier to find. Visit
http://www.usatoday.com/story/news/nation/2014/03/12/ejectable-recorders-plane-crash-data-voice-black-boxes/6338397/

It is not very easy to design an electable data recorder. I wish to stimulate a number of student inventors/designers to think about this challenge with this blog post, by discussing a number of issues involved.

The first question is about the conditions that should trigger ejection of the recorder. In a military aircraft, this could be when the pilot bails out using an ejection seat. Not in a civilian aircraft. Actual impact, particularly over water, might be too late a moment for the data recorder to eject; it might sucked deep into the water with the falling structures and get trapped under water.  A moment of high deceleration might indicate breaking up of the aircraft in mid-air. An impact will also cause a high level of deceleration.

Innovation requires trying out a variety of designs, without incurring a prohibitive cost. Integration of the device with the aircraft's systems will make these trials process very expensive. So, it might be worth trying out an externally mounted device not integrated with the avionics of the aircraft.  If this prevents cockpit voice recording, one can perhaps do without it to start with. Alternatively, one could try and have the cockpit voice signal sent to the recorder securely over an encrypted WiFi like transmission.

A possible option is that of a device that clings to the external surface of a plane and can fall off after ejecting, to float if it falls into a water body, and sends wireless signals irrespective of where it lands.

How would a device cling to the exterior of a plane when it has air flowing over it at a high speed? One option would be to use the power of the air rushing past the shell of the device to make the device cling tighter to the aircraft. A pitot tube mechanism could create a low pressure area inside the shell, making it cling tightly to the aircraft's surface. Plastic bolts or a magnetic device could provide for the primary attachment to the aircraft, with the partial vacuum created by the moving air providing an additional mechanism to bind the device to the aircraft. Ejecting the device would involve opening a valve that depletes the vacuum. It might also require a mechanical device that cuts or breaks the plastic bolt when necessary.

An external device can easily measure altitude as indicated by air pressure, as well as airspeed. Electronic accelerometers could add data about acceleration in three dimensions, and GPS coordinates as a function of time.

To save battery life, the ejected device may send wireless signals relatively infrequently. It could be designed to respond to polling signals sent by search aircraft or ships, like a transponder. The device might also have built-in solar cells to keep sending signals over weeks or months.

Given all the above-mentioned possibilities, it seems the time is ripe to launch several design projects. A design competition, with an attractive price, may be a good idea.

Having an aircraft sending data more or less continuously is not a bad solution. However, constant search for alternative solutions is necessary till all aircraft are fitted with affordable data recording devices.

Lastly, automobiles can also benefit from data recorders, even if they are not ejectable. Perhaps insurance companies should drive their use by offering lower insurance costs to cars fitted with working and regularly monitored data recorders. Automobile data recorders would have a moderating effect on drivers and make them less risk-prone.

Srinivasan Ramani

Thursday, March 13, 2014

Richard E. Merwin Student Scholarship


IEEE Computer Society is offering $40,000 in student scholarships from $1,000 and up to recognize and reward active student volunteer leaders in student branches or chapters who show promise in their academic and professional efforts. For additional information, visit

http://www.computer.org/portal/web/studentactivities/merwin?lf1=417765105e920816014288d18427682

Deadline for application is April 30, 2014.

Srinivasan Ramani

Sunday, February 23, 2014

Aids for the Visually Handicapped


IEEE Region 10 Humanitarian Technology Conference 2014 (R10 HTC) is to be held during August 2014, in Chennai, India. http://www.ieeer10htc.org This offers students an opportunity to carry out innovative projects which could have a social impact. The conference has a broad focus, but I will discuss one specific sub-area here, as I have explored this area to some extent.

Technology for the visually challenged
Look at a very simple device that sells for 30 British Pounds. Visit http://www.britishpathe.com/video/ultrasonic-torch   It seems to have been publicized from 1968. It must have benefitted millions of users. I looked for research on user experience and found a very nice paper from the same era, brimming with ideas. Visit http://www.rehab.research.va.gov/jour/68/5/1/91.pdf

This ultrasonic torch shows how valuable a simple device can be. In addition to giving an idea of obstacles in one’s path, it can even give some indication of the nature of the surface ahead.
Of course, there are a number of researchers working in this area, as illustrated by a couple of relatively recent papers: 
http://www.sersc.org/journals/IJDRBC/vol2/1.pdf

This area is rich for technical exploration and invention. You could use auto-focus mechanisms exploiting infra-red light, which work in broad daylight. Considering the large volume of production, I would expect necessary components to be commercially available at reasonable prices.

Cell phones have a lot of the infrastructure required to create an ultra-sound beam, to pick up the reflections and process them. Suitable additions to a phone can make implementation easy.
Signal processing techniques enable you to vary the frequency of the signal carried by the beam. This can enable sophisticated probing of the obstacles ahead of a visually challenged person.

It is not pure technology
A lot of technologists learnt a good deal from Steve Jobs work! He made it painfully clear to them that they will fail in the absence of respect for good design ideas and concern for the user experience. He taught that we should not make a device a mere tool, but something that it is a delight to own and use. This point of view is particularly important when we design anything for a visually challenged person.

It is also a matter of business
Innovation cannot stop with ideas, patents and publication.
The Wikipedia says that “Term of every patent in India is 20 years from the date of filing of patent application, irrespective of whether it is filed with provisional or complete specification”. Visit http://en.wikipedia.org/wiki/Indian_Patent_Office
The patent system provides a limited time to the inventor to “work” his invention and get a benefit; in return it compels him to disclose his invention for public good. 

Do we have entrepreneurs carefully watching for patents that expire? Do they consider manufacture and sale of devices covered by such patents? Do potential entrepreneurs among students look for these opportunities? No one stops an entrepreneur from innovating and inventing. One can always patent a new invention which vastly improves an old device covered by an expired patent. This gives a double strength – a tried and tested idea no longer covered by a patent and a new idea of one’s own covered by a new patent.

A Challenge
Let me conclude by proposing a challenge that may attract some of you. Can we manufacture and sell a device to help the visually challenged at the price of a reasonably good low-end cell-phone? I have in mind something that would cost about 2,500 Indian rupees (say 40 US dollars).
 
Srinivasan Ramani

Saturday, January 04, 2014

Cellular Technology cannot help Disaster Victims Outside a Cell Phone Tower’s Reach

  
It is amazing that a terrorist can buy a satellite phone. He can even rent it in certain countries. I understand that a week’s rental is about a thousand US dollars. However, a peaceful citizen caught in an accident can die because he or she cannot call for help. A fisherman caught in hurricane's aftermath can drift for days on the high seas without being able to seek help. A village facing a disaster can suffer for a week because its communication channels are cut. Such a problem did occur last year when floods hit Himachal Pradesh. 

I would like engineering students to ask themselves this technical question: Can we find out a way to ensure that affordable cell phones can send an SMS to some disaster management agency from anywhere in India?  Of course this may require addition of some hardware and software to the cell phone itself. It may require legal and operational safeguards to ensure that some idiot does not send an SMS as a practical joke, making people run around for nothing. At the moment let us just focus on technology. The practical problems mentioned can be solved.

A satellite which can pick up a weak signal and deliver it in a readable form to a ground station could be part of the solution. Telecom guys ought to ask what it takes to receive a 160 Byte signal from a handheld battery operated device and send it down. The principle of complexity inversion says that it may be cost effective to put the sophistication into the satellite or into the ground station rather than into millions of cell phones. The trick is to be able add no more than a couple of thousand rupees to the cost of the cell phone itself, even if it means that we put millions of rupees worth of equipment into the satellite or into each ground station.

I don’t think we ought to limit ourselves to 1 or 1.5 W transmissions from a cell phone. We could use well known techniques to use only a cell phone battery and still spend 10 W for 100 milliseconds. The signal does not have to be sent once only; we can have the phone repeat it a hundred times in a day.  
You can be sure that the “authorities” would not want you to escape their surveillance. No problem, they can tap the ground station for all they want.

Cell phone operators will want their pound of flesh. No problem – we can let them set up the ground stations and charge us a hundred times the standard rate for delivering an emergency SMS.

Are there solutions that do not involve satellites? Is there some part of the wireless spectrum in which a MHz channel be set apart for some form of transmission that will travel hundred kilometers well enough? Short wave? VHF with meteor trail reflection? Artificial ion trails created by simple rockets being fired periodically, say every hour on the hour?

Does the radiation of an emergency signal from a cell phone have to be omni-directional? Can’t we have some indication from the cell phone as to where we should point it so that a signal being sent out in a cone of (say) 60 degrees beam-width has a chance hitting the receiver or the relevant reflecting surface? If there is an ion trail involved, can the cell phone get a signal through it to create an audible alert?   

I am sure that we are smart enough to create a system that a human cry for help can be made to reach us, irrespective of where it originates! It can also give us the location code to tell us its origin. It is the job of engineers to find out how to do all this!


Srinivasan Ramani