What Hinders The Growth Of Robotics In India

Apollyon

Führer
Senior Member
Joined
Nov 13, 2011
Messages
3,134
Likes
4,573
Country flag
---- startups. If you want a secure future work hard and get into some PSU. :p
PS: Python is good for beginners or I say the early learners. For serious work C# and C++ is the real deal.

Edit: What's the fuss about IIT'ians in this Thread :confused:.
 
Last edited:

LurkerBaba

Super Mod
Joined
Jul 2, 2010
Messages
7,882
Likes
8,125
Country flag
---- startups. If you want a secure future work hard and get into some PSU. :p
PS: Python is good for beginners or I say the early learners. For serious work C# and C++ is the real deal.
C# is more or less similar to Java, so performance is not on par with C/C++

C, C++ and Objective-C have similar performance characteristics.
 

Apollyon

Führer
Senior Member
Joined
Nov 13, 2011
Messages
3,134
Likes
4,573
Country flag
C# is more or less similar to Java, so performance is not on par with C/C++

C, C++ and Objective-C have similar performance characteristics.
The only coding language I'm familiar with now is C# . I haven't used C, C++ and Python for a long time now.
I was never really interested in Coding and just switched on to different coding languages (C, Python, G, Java and C#) on need by need basis. :lol:
 

Peter

Pratik Maitra
Senior Member
Joined
Mar 3, 2014
Messages
2,938
Likes
3,341
Country flag
The only coding language I'm familiar with now is C# . I haven't used C, C++ and Python for a long time now.
I was never really interested in Coding and just switched on to different coding languages (C, Python, G, Java and C#) on need by need basis. :lol:
Oh that`s nice. Well I know only one language properly Java and C and C++ improperly. (I never prepared for those sems properly:rofl:)
 
Last edited:

Energon

DFI stars
Ambassador
Joined
Jun 3, 2009
Messages
1,199
Likes
767
Country flag
ROFL

Indian 'success' in IT are zero-innovation IT service companies. Real innovation requires top-notch engineering talent along with truckloads of money
Yup. I think the media spin on this topic has seriously obscured the actual value of the IT industry in India. And you're right it is not by any means an innovation oriented industry.

LurkerBaba said:
IMO joining BHEL, DRDO is a waste. It's better to start your own company or join a startup now that VC funding has started trickling in.

@Energon Check out this Delhi based startup - Grey Orange Robotics. They got funding from Tiger Global and Blume:

After Flipkart, Tiger Global bets on Indian robot maker Grey Orange Robotics - The Economic Times
In some countries government defense programs inadvertently make notable contributions to overall industrial growth. Senor and Singer for instance have shown how mandatory service in the IDF has contributed significantly to Israel's tech start up culture. India however is not one of those countries. And I absolutely agree with you in saying that governmental behemoths like DRDO and BHEL are a waste. If anything they are part of the problem because they've been hogging all the resources for decades without making significant contributions to overall industrial advancement.
In India's case promoting start ups is probably the best option. Grey Orange Robotics seems very impressive. I think their fortunes would have changed drastically if the retail sector had been opened up.
Overall though the success stories are still few and far in between, and as I mentioned earlier I honestly don't think India will truly industrialize unless the establishment gets its act together.
 
Last edited by a moderator:

boris

Regular Member
Joined
May 2, 2012
Messages
347
Likes
256
Country flag
ROFL

Indian 'success' in IT are zero-innovation IT service companies. Real innovation requires top-notch engineering talent along with truckloads of money



C/C++ is the real deal. Any CPU-intensive algo is going to be written in C/C++ or at max Java.

There is nothing 'modern' about Python. What's good about it is that it's not verbose, you can build things quickly with it. However, it falls flat in performance. A garbage-collected interpreted language cannot hold a candle to compiled language with manual memory management like C/C++. All of Python's high performance components are written as C extensions

There are some modern high performance languages though (Go and Rust come to mind)

==
=======================================================================================

IMO joining BHEL, DRDO is a waste. It's better to start your own company or join a startup now that VC funding has started trickling in.

@Energon Check out this Delhi based startup - Grey Orange Robotics. They got funding from Tiger Global and Blume:

After Flipkart, Tiger Global bets on Indian robot maker Grey Orange Robotics - The Economic Times
Excellent answer. The problem with most Indian colleges is that firstly both C and C++ are not taught correctly, I mean students are made to write the code in a very outdated Turbo-C IDE which was made even before C++ was standardized, hence many students tend to think they are not so powerful but it isn't so even all these decades later C is still de-facto the best language for systems programming. All OS Kernels are written in C , most database components are either written in C and C++ except for the newer NoSQL databases coming out, browsers,compilers,game engines,embedded systems,search engine algorithms etc you name it are either written in C or C++. Performance wise as of now these two languages are unbeatable.

I sort of agree on your statement as far as Python is concerned, it is easy to learn, has a truckload of API's to use for data science(scipy,numpy),NLP(NLTK) and stuff is fast to create but the language is really slow compared to most languages but still it is a good choice now
especially with newer platforms like Django etc.
 
Last edited by a moderator:

Dovah

Untermensch
Senior Member
Joined
May 23, 2011
Messages
5,614
Likes
6,793
Country flag
C# is more or less similar to Java, so performance is not on par with C/C++

C, C++ and Objective-C have similar performance characteristics.
Java is suited for enterprise applications where import overhead does not matter that much and business implementation is more important than memory economics.

C++/C as you correctly said is more CPU intensive, more for product based industries. Assembly is where it's at though. :D
 

boris

Regular Member
Joined
May 2, 2012
Messages
347
Likes
256
Country flag
Well C is a bit outdated. You may say Java is the best compared to everything. There are many problems in C. One that immediately comes to my mind is that of array size in C and Java. Also remember the trouble one will have in strings with C. IMO Java is best(not bcoz I am more suited to Java lol).

All I was wanting to say was that python is the latest craze among programmers of this age and Indian cse students should have a basic knowledge about it.
Are you talking about fixed array size? well that can be easily solved via pointers by allocating memory on the heap rather than the stack using functions like malloc() and then freeing the memory using free() when done(its C so no GC), in C++ mostly it is better to use vectors instead of arrays in contrast to C, C++ has STL for readymade data-structures and an excellent string library.

C and C++ are different languages once your project reaches some level of complexity, check out open source code for libraries for eg : a Data Structures library the C and C++ version is very different , the C++ version will heavily use templates and concepts like template metaprogramming. C++ shares this similarity with C that its architecture-wise backward compatible with C otherwise its quite different and has become even more amazing with C++11 standard that has come out.

and another thing C is quite powerful although like Java and C++ it does not offer object-orientation by default, a programmer can write OOP like code in C.
 

boris

Regular Member
Joined
May 2, 2012
Messages
347
Likes
256
Country flag
Saar, there are some major differences between Java and C/C++. Java is an interpreted language, while the latter is compiled, no matter how much you optimize you'll never get the same performance. Sun's HotSpot JIT (Just In Time compiler) is pretty good though. Another difference is memory management. Pointers vs Garbage Collection, Pointers win no matter how annoying they are.

Complex computer games, device drivers, large applications, sophisticated algorithms or basically anything which requires high performance WILL be written in C/C++.

Python is pretty good in certain cases and very useful to just learn programming.



As I said, make use of VC funding or join an existing startup. BHEL, DRDO means complete waste of talent. It's better to immigrate to American where some $$ will flow into India via remittance :p
Exactly, although its fun to note that HotSpot is written in C++. Yeah many Java lovers love to jump on the GC point but in C++ using RAII,smart pointers especially with Boost pretty much eliminates the need for GC. Pointers maybe annoying but they are darn powerful.

Anyway in the US I am seeing many Java gurus migrate to Scala.
 

LurkerBaba

Super Mod
Joined
Jul 2, 2010
Messages
7,882
Likes
8,125
Country flag
OS Kernels are written in C , most database components are either written in C and C++ except for the newer NoSQL databases coming out, browsers,compilers,game engines,embedded systems,search engine algorithms etc you name it are either written in C or C++. Performance wise as of now these two languages are unbeatable.
MongoDB is written in C++ :p

Java is suited for enterprise applications where import overhead does not matter that much and business implementation is more important than memory economics.
Saar you sound like a product manager

Exactly, although its fun to note that HotSpot is written in C++. Yeah many Java lovers love to jump on the GC point but in C++ using RAII,smart pointers especially with Boost pretty much eliminates the need for GC. Pointers maybe annoying but they are darn powerful.

Anyway in the US I am seeing many Java gurus migrate to Scala.
Serious folk dependent on the JVM are trying out Scala and Groovy, hipsters are using Clojure

@LurkerBaba, So what are you saying? Indian IT sector has no innovation to speak of??
Indian IT services sector (TCS, Infy, Wipro and the likes) does zero innovation.
 
Last edited by a moderator:

boris

Regular Member
Joined
May 2, 2012
Messages
347
Likes
256
Country flag
MongoDB is written in C++ :p



Saar you sound like a product manager



Serious folk dependent on the JVM are trying out Scala and Groovy, hipsters are using Clojure



Indian IT services sector (TCS, Infy, Wipro and the likes) does zero innovation.
Oh yeah all the hipsters are promoting Clojure as the next big thing for writing Machine Learning algo's, a lot of them suggested me that instead of Scala whenever I want to try out functional programming. It reminded me of Lisp/Scheme that I read when I tried reading "Structure and Interpretation of Computer Programs"(one book that has become sort of a challenge to read, along with Don Knuth's books :lol: :lol:) I decided Scala is better initially , I mean if I really wanted to take the FP challenge I would have ignored all choices and jumped off to Haskell.
 

no smoking

Senior Member
Joined
Aug 14, 2009
Messages
5,000
Likes
2,302
Country flag
Technology does not.... the industrial model.
...................

of misplaced socialist ideals.
The best understanding of industrilization from an Indian so far, but still there are something not quite right.

For example:

A study of industrialization starting with Western Europe, North America, Japan, S Korea and now China clearly shows that it takes decades to accomplish this task and all steps have to take place simultaneously.
No, you can make them happen simultaneously, industrilization requires certain sequence. You will have to find the one which suits you otherwise you will find youself run out of money and resources very soon.

In India's case it is this simultaneous development that has been the problem IMO. Everything is fine in bits and pieces- good universities, ample funding from government, talented private sector operative etc. but somehow it just doesn't come together.
No, everything won't be fine for any country which try to industrilize itself in 20 century. It is the same for india.
India has good universities, but very few;
Ample funding from government? Far from it! The probem with any developing country is: they don't have money for the R&D in EVERY industry department.
Private sector? Trust me, there is no private company would be interested in those basic R&D which won't see any profit in the decades.

Over the past few decades China has been investing heavily into establishing the pyramid base (primarily via industrial espionage) and establishing sound industrial practices.
No, pyramid base can't be built on industrial espionage. The espinage can provide you the final anwer, but it can't give the knowledge of every single step. You have to learn it by yourself. For example, your spy can get the map of most advanced fighter jet engine, but you can't simply build a new engine based on it. Why? You don't have the same parts and material unless the original producer is willing to offer them. So, you have to use your own material or parts. But soon you will find it doesn't work because some parameters of your materials are different, so you have to change the structure of the engine, then you will find new problems caused by this change. Every single variance will bring in a lot new problems requiring reseach work in different depart. Finally, you will find you are actually working on a completely new engine---your own engine.


I don't think technological progress will take hold in India until the government invests in a robust infrastructure and simultaneously removes all the barriers they have erected through decades of misplaced socialist ideals.
Based on Chinese experience, that is just the beginning of a painful process.
 

pmaitra

Senior Member
Joined
Mar 10, 2009
Messages
33,262
Likes
19,593
Private sector? Trust me, there is no private company would be interested in those basic R&D which won't see any profit in the decades..
Many people don't realize that high tech defense related innovation can only be done by government companies. Private companies are only good when it comes to making profits, and they have little room for failure. However, in research, failure is the path to success. Private companies do not allow that opportunity, because, the realities in which private companies operate are about immediate profits.
 

Energon

DFI stars
Ambassador
Joined
Jun 3, 2009
Messages
1,199
Likes
767
Country flag
No, you can make them happen simultaneously, industrilization requires certain sequence. You will have to find the one which suits you otherwise you will find youself run out of money and resources very soon.
No, everything won't be fine for any country which try to industrilize itself in 20 century. It is the same for india.
India has good universities, but very few;
Ample funding from government? Far from it! The probem with any developing country is: they don't have money for the R&D in EVERY industry department.
Private sector? Trust me, there is no private company would be interested in those basic R&D which won't see any profit in the decades.
The coordination I'm referring to is in regards to government, academic institutions and industry; and yes as far as India is concerned it has to be simultaneous. China is unique (in the current cohort) in that all three aspects are controlled by the government.
Based on my observations of India it seems to me that the individual components have developed somewhat over time. What hasn't developed however is the matrix linking the different components in order to make them synergistic. Yes, the scale is certainly inadequate (the small number of good universities etc.), however if there was an appropriate supportive system in India all the individual components can be expanded over a relatively short period of time.
In terms of funding the Indian establishment certainly has a decent amount of money to invest (albeit not as much as developed industrialized nations or China). The funding of science is a science in of itself; in some ways it can be considered an infrastructural feature akin to taxation system. The funding mechanisms in India are woefully inept. However this is a chicken and egg conundrum because without a reliable research infrastructure any potential investments are bound to produce low yields making them unpalatable.


no smoking said:
No, pyramid base can't be built on industrial espionage. The espinage can provide you the final anwer, but it can't give the knowledge of every single step. You have to learn it by yourself. For example, your spy can get the map of most advanced fighter jet engine, but you can't simply build a new engine based on it. Why? You don't have the same parts and material unless the original producer is willing to offer them. So, you have to use your own material or parts. But soon you will find it doesn't work because some parameters of your materials are different, so you have to change the structure of the engine, then you will find new problems caused by this change. Every single variance will bring in a lot new problems requiring reseach work in different depart. Finally, you will find you are actually working on a completely new engine---your own engine.
I truly hope you're saying this out of some obligatory keyboard nationalism. Because if there's even the slightest chance you're a professional researcher then we'd both know you're full of $h!t and that would make things very awkward. Everyone who actually knows what they're talking about is well aware that the overwhelming bulk of the theft does not include fighter jet engines and defense tech but rather oodles and oodles of foundational science research in the form of technical SOP manuals, individual wet lab experiments etc. etc. I have colleagues who have seen all their experiments stolen and aped. Within the last 4 years I haven't been to a single conference where the topic of theft wasn't a main topic of discussion. Most of my collaborators are from industry and their stories are even worse. And any project with the word "collaboration" in it that involves China is pretty much a lost cause. It doesn't matter if it's a windmill farm or the international thermonuclear experimental reactor, the result is the same; the Chinese will be running a parallel program based on all the information stolen from other members. And industrial MoUs"¦ LOL if a Chinese partner is involved you might as well wipe you're a$$ with those. I work at a tier 1 research university and two years ago the experimental data from the freshman undergraduate lab computers was stolen by some hacker(s) from China"¦ seriously freshman undergrads"¦ which means there's some azhole in China dropping ice cubes into a goldfish bowl making sure the goldfish's respiration rate decreases (spoiler alert: it does). Heck I know a guy who went to a fashion school that busted a ring of Chinese foreign students who were making Xerox copies of fabric patterns and faxing them to China late at night"¦ yea I didn't even know that was a thing"¦ but it is. So I don't know what the CCP party line is these days but China's industrial knowledge base is almost entirely founded on blatant theft.
But in all fairness this is not a China specific issue (albeit the scale is unprecedented) industrial espionage has been around ever since industrialization first started.
 

no smoking

Senior Member
Joined
Aug 14, 2009
Messages
5,000
Likes
2,302
Country flag
The coordination I'm referring to is in regards to government, academic institutions and industry; and yes as far as India is concerned it has to be simultaneous. China is unique (in the current cohort) in that all three aspects are controlled by the government.
It is not about whether or not the gov is in control of everything. For any developing country, we all face the same problem at the begining stage: lack of resources and funds. China got over this stage in 1990s while India and other developing countries still stuck in there. What make China, korea and Japan succeeded in last 100 years? We didn't do things simultaneously. Each stage has its own mission and goal.



Based on my observations of India it seems to me that the individual components have developed somewhat over time. What hasn't developed however is the matrix linking the different components in order to make them synergistic. Yes, the scale is certainly inadequate (the small number of good universities etc.), however if there was an appropriate supportive system in India all the individual components can be expanded over a relatively short period of time.
No, your individual components are slightly better than those south eastern Asians: they sell more products only because they have a bigger domestic market. They didn't do much basis R&D: spend hundredes millions dollars on a single material and then lock it in the safe for 20 years until market matures?

In terms of funding the Indian establishment certainly has a decent amount of money to invest (albeit not as much as developed industrialized nations or China). The funding of science is a science in of itself; in some ways it can be considered an infrastructural feature akin to taxation system. The funding mechanisms in India are woefully inept. However this is a chicken and egg conundrum because without a reliable research infrastructure any potential investments are bound to produce low yields making them unpalatable.
No, in india, the fund demand from gov would be far bigger than China or other industrilized countries! For example, if we build an lab of one industry sector which costs 100billion dollars. Western countries had 100s years to make that investment, Japanese--60 years, Chinese--40 yeras. The more years you got, the less amount you will need annually.


I truly hope you're saying this out of some obligatory keyboard nationalism. Because if there's even the slightest chance you're a professional researcher then we'd both know you're full of $h!t and that would make things very awkward. Everyone who actually knows what they're talking about is well aware that the overwhelming bulk of the theft does not include fighter jet engines and defense tech but rather oodles and oodles of foundational science research in the form of technical SOP manuals, individual wet lab experiments etc. etc. I have colleagues who have seen all their experiments stolen and aped. Within the last 4 years I haven't been to a single conference where the topic of theft wasn't a main topic of discussion. Most of my collaborators are from industry and their stories are even worse. And any project with the word "collaboration" in it that involves China is pretty much a lost cause. It doesn't matter if it's a windmill farm or the international thermonuclear experimental reactor, the result is the same; the Chinese will be running a parallel program based on all the information stolen from other members. And industrial MoUs"¦ LOL if a Chinese partner is involved you might as well wipe you're a$$ with those. I work at a tier 1 research university and two years ago the experimental data from the freshman undergraduate lab computers was stolen by some hacker(s) from China"¦ seriously freshman undergrads"¦ which means there's some azhole in China dropping ice cubes into a goldfish bowl making sure the goldfish's respiration rate decreases (spoiler alert: it does). Heck I know a guy who went to a fashion school that busted a ring of Chinese foreign students who were making Xerox copies of fabric patterns and faxing them to China late at night"¦ yea I didn't even know that was a thing"¦ but it is. So I don't know what the CCP party line is these days but China's industrial knowledge base is almost entirely founded on blatant theft.
Relax, I didn't deny the chinese technology theft activities! OK?
All I pointed out is that you can't build up your R&D basis with this kind of espionage. Unless you already have a decent science and industrial basis, these tech espionage won't help you much! And this basis, my friend, it is the thing that make China unique among developing countries. Another tip is you have to make this basis ready before you even think about stealing or buying.

But in all fairness this is not a China specific issue (albeit the scale is unprecedented) industrial espionage has been around ever since industrialization first started.
Of course, China is trying to build an industrilized world for 1.4 billion population which is almost the total industrilized population in the developed world. That means China need every single industry department!
 

Latest Replies

Global Defence

New threads

Articles

Top