Reverse Engineering Is Extremely Difficult

[Martian2]

For those of you who think reverse-engineering is merely obtaining a sample of high-technology and pushing the "copy" button, think again. To produce the Yu-6 torpedo, a reverse-engineered Mark 48 torpedo, it took ten years and 18 patents for brilliant teams of Chinese scientists to reproduce the Mark 48 torpedo.

Reverse-engineering is not copying. Reverse-engineering is the reinvention of the entire manufacturing process, including patents, to build the high-technology weapon or equipment. Unless you have dedicated brilliant scientists and a sufficiently-advanced high-technology base, reverse-engineering of high-technology weapons or equipment is not possible.

Let's examine the excruciating process to reverse-engineer a mere torpedo. Please read the following paragraphs on the reverse-engineering of the Mark 48 torpedo. When you're done, tell me that you're not impressed by the scientists in a developing country that reinvented a high-technology weapon made by the world's leading technological power.

Yu-6 torpedo - Wikipedia, the free encyclopedia

"Yu-6 torpedo
From Wikipedia, the free encyclopedia

Yu-6 (鱼-6) torpedo is the Chinese equivalent of the Mark 48 torpedo. In addition to wire and active / passive homing guidance, wake homing guidance is also incorporated. Many domestic Chinese sources have claimed that Yu-6 torpedo is in the same class as the Mk 48 Mod. 4 torpedo, but official information of Yu-6 torpedo released by the Chinese government is limited and such claim thus cannot be confirmed by sources outside China.

Background

At least one Mark 48 torpedo was reportedly recovered by Chinese fishermen in the late 1970s or early 1980s, and China might have begun the reverse engineering in the 1980s. However, due to the inexperience of the Chinese technological base at the time, as well as the concentration on economic development, most of the reverse engineering attempt was put on hold after research had been completed on Otto fuel II, wire guidance and some other subsystems, but some research continued on much smaller scale. The Yu-6 torpedo developmental program experienced a revival when the Chinese military realized that despite developing several torpedoes including the Yu-1, Yu-2, Yu-3, Yu-4, and Yu-5, the obsolete doctrine of having separate ASuW and ASW torpedoes proved to be unsuited for modern naval warfare and the Chinese navy needed a torpedo for both ASuW and ASW. As a result, the Yu-6 program was fully resumed in 1995 and 705th Institute was named as the primary contractor, Mr. Dong Chunpeng (董春&#40527 as the general designer.

Development

One of the difficulties encountered was that the Yu-6 torpedo had greater operating depth than all previous Chinese torpedoes, and China could not build the new casing needed. A brand new alloy was required to cast the outer casing of the Yu-6 torpedo, and under the leadership of Professor. Ding Wenjiang (丁文&#27743 of material science at Shanghai Jiao Tong University, the problem was solved when ZLJD-1S alloy was successfully developed and used to cast the casing for Yu-6 torpedo. Professor He Yuyao (贺昱&#26332 was in charge of developing power module for the newly developed computer of Yu-6 torpedo, and due to extremely high standard of the processing capability that demanded very advanced power source, this power module took three years to complete, (1999 – 2001).

Another huge obstacle faced was the casing for the acoustic seeker required new materials to build, but China had never had any experience in this field when the reverse engineering was first attempted. Despite most reverse engineering attempt was put on hold, research in this field continued and Tianjin Rubber Research Institute (also known as Tianjin Municipal Rubber Industry Research Institute (天津市橡胶工业研究&#25152 was tasked to develop the rubber needed for the acoustic seeker casing. A team of 7 scientists including Shen Yingjun (申英&#20426, Hou Yehua (侯月&#21326, Zhang Jianguo (张建&#22269, Ma Gangying (马刚&#33521, Zhang Lixia (张立&#20384, An Jiazhu (安家&#26609 and Zhang Suqin (章素&#29748 was formed, and eventually developed new rubber and production technique needed October, 1994, shortly before the Yu-6 program was fully resumed in 1995. The resulting rubber not only met the original requirement, but also exceeded it, with density reaching 1.098 ton per cubic meter, speed of sound reaching 1,551 metre per second, and surface roughness reaching 0.16 micrometer.

The propulsion system was the biggest obstacle in the development of Yu-6 torpedo. A team of three scientists including Ms. Su Li (苏&#20029, Mr. Mao Yuanfu (毛元&#31119 and Mr. Wang Lisong (王立&#26494 of Harbin Electro Carbon Research Institute (哈尔滨电碳研究&#25152 was tasked to develop the graphite material used to make engine valves and other components. By September, 1998 the new graphite material, designated as M130, had been successfully developed and then utilized for Yu-6 torpedo. The piston ring of the engine was developed successfully in December 2003 by Yizheng Shuanghuan Piston Ring Co. Ltd. (仪征双环活塞环有限公&#21496. Mr. Wang Guozhi (王国&#27835 was in charge of noise reduction and his successful research in this field resulted in him winning a second place in the Chinese national scientific and technology advancement award in 1998.

Over two-thirds of the technologies used for Yu-6 torpedo were new to the Chinese, and there were some serious doubts that China could complete the project on its own. Mr. Dong Chunpeng (董春&#40527, a University of Science & Technology of China graduate in 1966 who then worked for 705th Institute since was determined to succeed, which he did after developing 18 patents including four in fields that China had never had any experience. After ten years of development, Yu-6 torpedo was finally accepted into service in 2005, and for the success, Mr. Dong Chunpeng (董春&#40527 was awarded the 2006 Scientific and Technological Advancement Award at the end of February in Beijing. One characteristic of Yu-6 torpedo is its high performance processor. In comparison to the Motorola 68000 or Intel 8086 microprocessors commonly used on most western torpedoes, the microprocessor used for Yu-6 torpedo is at least equal to the Intel 80486 class. Some domestic Chinese sources have claimed that Loongson-1 is used for Yu-6 torpedo, and the operating system is the Kylin operating system developed by the National Defense Science and Technology University, but this has yet to be confirmed by both the Chinese government and sources outside China. Another characteristic of Yu-6 torpedo is that the transducer array of its acoustic seeker has at least 55 transducers, more than that of US Mark 48 torpedo, which has a total of 52 transducers, but the exact number has not been released. Yu-6 torpedo was also the first Chinese torpedo designed with the concepts of modular design and open architecture software programming in mind, so that when new technologies and programs become available, they could be readily incorporated. In comparison to the first generation Chinese wire guided torpedo Yu-5 torpedo, which must use acoustic guidance in the terminal stage or when the wire is severed, the wire and acoustic guidance can be switched from one to another at any time in the Yu-6 torpedo. Furthermore, when the wire of Yu-6 torpedo is severed, the targeting information stored in the memory would enable the computer onboard to calculate the approximate new location of the target, augmenting the acoustic homing to achieve a higher kill probability.

Specifications

* Diameter: 533 mm
* Guidance: passive / active acoustic homing + wake homing + wire guidance
* Propulsion: Otto fuel II
* Speed: maximum > 65 kt (for attack)
* Range: maximum 45 km+ (at the cruise speed)"

 

[new wave]

Bro, good post, but i am afraid it will turn into a troll thread pretty soon, i hope i am wrong, appreciate your efforts though.

 

[jagjitnatt]

Well reverse engineering is not easy but inventing something from scratch is very tough.

When reverse engineering, one knows what he's trying to build, he can understand how things work and what technology he is working on, whereas when building something from scratch, the inventor needs to try a million things, experiment, find the best route, check the compatibility, test the product for reliability and performance issues.

Reverse engineering is like walking on a treaded path whereas inventing something is creating your own path.

There's a huge difference. So relatively reverse engineering is too easy as compared to building something from scratch.

 

[Martian2]

Well reverse engineering is not easy but inventing something from scratch is very tough.

When reverse engineering, one knows what he's trying to build, he can understand how things work and what technology he is working on, whereas when building something from scratch, the inventor needs to try a million things, experiment, find the best route, check the compatibility, test the product for reliability and performance issues.

Reverse engineering is like walking on a treaded path whereas inventing something is creating your own path.

There's a huge difference. So relatively reverse engineering is too easy as compared to building something from scratch.

Your definition of "too easy" seems to include ten years, "18 patents including four in fields that China had never had any experience," "another huge obstacle faced was the casing for the acoustic seeker required new materials to build, but China had never had any experience in this field," and "over two-thirds of the technologies used for Yu-6 torpedo were new to the Chinese."

You have a very strange definition of "easy." You and I seem to have learned very different English. I would not call the tremendous Chinese program as effortless.

Easy | Define Easy at Dictionary.com

"eas·y /ˈizi/ Show Spelled [ee-zee] Show IPA adjective,eas·i·er, eas·i·est, adverb, noun
–adjective

1. not hard or difficult; requiring no great labor or effort: a book that is easy to read; an easy victory.
...
9. effortlessly clear and fluent: an easy style of writing."

 

[below_freezing]

only a non-scientist or non-engineer can say that. reverse engineering is extremely difficult because there are limitations to your own ability to manufacture parts. what if there is a piece that has to be built to 0.0001 cm precision or it doesn't work, but your best machine tools can only build to 0.1 cm precision? you can either buy a new machine (like SK does), give up and buy the whole thing (some large country to our south), or redesign that part so it doesn't have to be that precise. but if you redesign that part, you have to redesign everything that part involves, etc etc...

 

[jagjitnatt]

Your definition of "too easy" seems to include ten years, "18 patents including four in fields that China had never had any experience," "another huge obstacle faced was the casing for the acoustic seeker required new materials to build, but China had never had any experience in this field," and "over two-thirds of the technologies used for Yu-6 torpedo were new to the Chinese."

You have a very strange definition of "easy." You and I seem to have learned very different English. I would not call the tremendous Chinese program as effortless.

Easy | Define Easy at Dictionary.com

"eas·y /ˈizi/ Show Spelled [ee-zee] Show IPA adjective,eas·i·er, eas·i·est, adverb, noun
–adjective

1. not hard or difficult; requiring no great labor or effort: a book that is easy to read; an easy victory.
...
9. effortlessly clear and fluent: an easy style of writing."

You forgot to refer to dictionary for the word RELATIVE. Go figure that first.

 

[sohan]

Reverse engineering is hard, innovation is far far harder.

 

[Martian2]

You forgot to refer to dictionary for the word RELATIVE. Go figure that first.

I don't think you understand my article. The point of my article is that reverse-engineering is reinvention and not copying. I have never made the ridiculous claim that reverse-engineering is harder than original invention. The point is that reverse-engineering is a herculean effort, especially by a developing country, that requires ten years, many patents, and exploring new fields of science.

Your first comment about reverse-engineering being relatively "too easy" is irrelevant and does not pertain to the main idea regarding the difficulty of reverse-engineering.

 

[Martian2]

Reverse engineering is hard, innovation is far far harder.

That is precisely my point. Cutting-edge innovation proceeds very slowly. However, especially for a developing country, "reverse engineering is hard."

 

[Martian2]

Oh Martian2, here you go again... ****. Stop communist ****ing on this forum for a petty 50 cent coin. Oh wait, did that go up to 70 a while ago? Moreover, you're using wikipedia. A ******* 10 year old can edit that ****.

1) Reverse engineering is an extremely useful skill to own, but by no means, something to be proud of and dignified for. It pisses me off reading some of your posts. This world is driven by invention and innovation, not reinvention and reverse engineering.

2) The ******* Loongson processor you're ******* about is only designed in Beijing, but fabricated by STM, an Italian-French company, which purchases its production lines from the States. You commies can't make **** chips out of nothing. You guys're lagging behind in microarchitectures while Intel's on nano.

Jagjitnatt, this is one of the few issues that I agree with and commend you Indians on.

Feel free to rant and ignore the main idea in the article, which is that reverse-engineering is very difficult and involves a complete reinvention of the technology. For example, the U.S. is the pioneer in radar-absorbent-material (i.e. RAM) coating. However, the process of reverse-engineering or reinventing your own RAM coating will be a long and difficult road.

I don't care if you rant at me. I just want the readers to understand that reverse-engineering is reinvention, not copying, and very difficult.

By the way, it is unrealistic to expect a developing country to innovate very much. There is a reason that they're called developing countries.

Look at the following list and tell me how far the developing countries are from the technological leader. Innovation is not realistic. As developing countries, the best that they can expect is reverse-engineering.

"For 2009, China passed Italy to claim the ninth-highest rank for countries that receive the most patents in the United States.

http://www.uspto.gov/web/offices/ac/ido/oeip/taf/cst_all.htm

Patents granted by the United States for the year 2009.

1. U.S. 95,037 patents
2. Japan 38,066
(Greater China 10,638)
3. Germany 10,353
4. South Korea 9,566
5. Taiwan 7,781
6. Canada 4,393
7. U.K. 4,011
8. France 3,805
9. China 2,270
10. Italy 1,837
...
India 720
Hong Kong 587 (Patent office counts Hong Kong as a separate entity)
Singapore 493
Russian Federation 204
Brazil 148"

 

[Guest]

Well reverse engineering is not easy but inventing something from scratch is very tough.

When reverse engineering, one knows what he's trying to build, he can understand how things work and what technology he is working on, whereas when building something from scratch, the inventor needs to try a million things, experiment, find the best route, check the compatibility, test the product for reliability and performance issues.

Reverse engineering is like walking on a treaded path whereas inventing something is creating your own path.

There's a huge difference. So relatively reverse engineering is too easy as compared to building something from scratch.

understanding how things work does not mean you can get it works. take this torpedo for example, even if they know exactly how it works. they can not produced the components. to manufacture component A, production line A is needed, to produced comonent B which is needed in production lina A, production line B is needed. this circle will go on and on unitll everything can be produced.

so instead of the MK48, what chinese done was to duplicate the whole industry needed to produce the torpedo. however, when all these are done, they can begin to invent.

 

[SinoIndusFriendship]

Oh Martian2, here you go again... ****. Stop communist ****ing on this forum for a petty 50 cent coin. Oh wait, did that go up to 70 a while ago? Moreover, you're using wikipedia. A ******* 10 year old can edit that ****.

1) Reverse engineering is an extremely useful skill to own, but by no means, something to be proud of and dignified for. It pisses me off reading some of your posts. This world is driven by invention and innovation, not reinvention and reverse engineering.

2) The ******* Loongson processor you're ******* about is only designed in Beijing, but fabricated by STM, an Italian-French company, which purchases its production lines from the States. Designing chips and ******* manufacturing them is 2 wholly different things, you ****. You commies can't make **** chips out of nothing. You guys're lagging behind in microarchitectures while Intel's on nano. Stop getting all cocky over some Italian-French chip, you ****.

Jagjitnatt, this is one of the few issues that I agree with and commend you Indians on.

That is false. Any modern chip-fabrication plant can manufacture a CPU. The issue of being able to design a chip is fairly complex. Designing a new type (e.g. MIPS, Arm, PowerPC, x86, Sparc, etc) is more difficult, but not that more difficult. The main reason it is not done is because a whole entire slew of new software from firmware, compilers, kernel, OS, network stack, applications have to be developed, tested and debugged. This is an onerous task, much more onerous than the actual CPU. Hence, Loongson opted for the MIPS architecture by licensing it and the 3F serious possess x86 emulation to take advantage of the off-the-shelf software. STM is only one of a several companies involved because of licensing issues, NOT design.

Ideally, I would like to have the entire hardware-software ecosystem, from programming language, compiler, CPU, peripherals, kernel, OS, to applications being home-made. The path currently taken is parallel development, where many components are being developed simultaneously --- and when one is mature and stable enough it is used to replace the foreign component.

Take for example QNX (hope you've heard of them), it is a real-time OS company based out of Ottawa Canada. This is a very simplistic RTOS, yet for the past 10 years little improvement has been made. It is now purchased by the makers of Blackberry, RIM. Whether it will find market success is slim, and whether it will see further development is even more slim. The point here is not whether other countries have potential to develop a CPU (Japan for instance does), it also involves the drive and commitment to see it to fruition.

True in that this is one area that the USA has a large monopoly over. It is also true that no other nation has been able to challenge the USA on CPU & Software.... yet. That's the beauty. Yet.

 

[SinoIndusFriendship]

understanding how things work does not mean you can get it works. take this torpedo for example, even if they know exactly how it works. they can not produced the components. to manufacture component A, production line A is needed, to produced comonent B which is needed in production lina A, production line B is needed. this circle will go on and on unitll everything can be produced.

so instead of the MK48, what chinese done was to duplicate the whole industry needed to produce the torpedo. however, when all these are done, they can begin to invent.

Jag gets very frustrated when there's a thread or when someone seems to "congratulate Chinese on their accomplishments". This has less to do when whether the achievement deserve merit or not, but more on his own frustration at the lack of progress Bharat is making. It's called "selective cognition".

Take for example the following statement often made by Bharatis, "India is the 2nd fastest growing economy in the world". Is this true? Not exactly. Well they might argue it's the 2nd fastest major economy in the world. Is this true? No it isn't either. Just focusing on Asia alone, Vietnam is actually Asia's 2nd fastest growing economy after China. Now why didn't the Bharatis consider Vietnam? Perhaps it's because Vietnam isn't a major economy compared to India, they might argue. Well that isn't quite true either. If you discount Vietnam's economy as "insignificant" you might as well also discount Bharat's economy, because Vietnam's economic and industrial strength & size is nearly HALF OF INDIA'S. Yes, you heard correctly, nearly half of India's. The scientific and engineering power and achievements coming out of Vietnam often drawfs that of India. And Vietnam's potential as the newest East Asian Powerhouse is still just lurking around the corner. Keep in mind Vietnam's achievements are greatly dwarfed by China.

So this is the core essence of the cause for frustration felt by many Bharatis.

 

[Martian2]

The ten years, 18 patents, and new fields of material science and propulsion technologies to develop the Yu-6 torpedo (e.g. Mark 48-class torpedo) were very impressive. However, the Type 052C Aegis-class destroyer is unbelievable.


Yu-6 torpedo


Type 052C Aegis-class destroyer

 

[ao333]

Feel free to rant and ignore the main idea in the article, which is that reverse-engineering is very difficult and involves a complete reinvention of the technology. For example, the U.S. is the pioneer in radar-absorbent-material (i.e. RAM) coating. However, the process of reverse-engineering or reinventing your own RAM coating will be a long and difficult road.

I don't care if you rant at me. I just want the readers to understand that reverse-engineering is reinvention, not copying, and very difficult.

By the way, it is unrealistic to expect a developing country to innovate very much. There is a reason that they're called developing countries.

Look at the following list and tell me how far the developing countries are from the technological leader. Innovation is not realistic. As developing countries, the best that they can expect is reverse-engineering.

"For 2009, China passed Italy to claim the ninth-highest rank for countries that receive the most patents in the United States.

Patents By Country, State, and Year - All Patent Types (December 2009)

Patents granted by the United States for the year 2009.

1. U.S. 95,037 patents
2. Japan 38,066
(Greater China 10,638)
3. Germany 10,353
4. South Korea 9,566
5. Taiwan 7,781
6. Canada 4,393
7. U.K. 4,011
8. France 3,805
9. China 2,270
10. Italy 1,837
...
India 720
Hong Kong 587 (Patent office counts Hong Kong as a separate entity)
Singapore 493
Russian Federation 204
Brazil 148"

Sir, for the record, you can patent for a chair.

That is false. Any modern chip-fabrication plant can manufacture a CPU. The issue of being able to design a chip is fairly complex. Designing a new type (e.g. MIPS, Arm, PowerPC, x86, Sparc, etc) is more difficult, but not that more difficult. The main reason it is not done is because a whole entire slew of new software from firmware, compilers, kernel, OS, network stack, applications have to be developed, tested and debugged. This is an onerous task, much more onerous than the actual CPU. Hence, Loongson opted for the MIPS architecture by licensing it and the 3F serious possess x86 emulation to take advantage of the off-the-shelf software. STM is only one of a several companies involved because of licensing issues, NOT design.

Ideally, I would like to have the entire hardware-software ecosystem, from programming language, compiler, CPU, peripherals, kernel, OS, to applications being home-made. The path currently taken is parallel development, where many components are being developed simultaneously --- and when one is mature and stable enough it is used to replace the foreign component.

Take for example QNX (hope you've heard of them), it is a real-time OS company based out of Ottawa Canada. This is a very simplistic RTOS, yet for the past 10 years little improvement has been made. It is now purchased by the makers of Blackberry, RIM. Whether it will find market success is slim, and whether it will see further development is even more slim. The point here is not whether other countries have potential to develop a CPU (Japan for instance does), it also involves the drive and commitment to see it to fruition.

True in that this is one area that the USA has a large monopoly over. It is also true that no other nation has been able to challenge the USA on CPU & Software.... yet. That's the beauty. Yet.

And no, I'm not saying that designing is, but relatively, much easier than chip fabrication. AMD had 64 bit designs ready WAY before the plants could make the compatible hardware. Heck, they even have 128 bits ready now. How many arduous years of overtures will it take for STM to procure the lines capable of producing those?

I recently just heard of a scandel in Shanghai's Fudan university of their capabilities to produce x86 80486-comparable chips. Guess what? It turned out to be a 486 with the intel tradmark scrubbed off! Evidently, China does not have the capacity to produce 486 equivalent chips or ... err... even LCD boards, which you guys produce using patented Korean technology.

Martian2: I simply adore the 052C. But it's just too highly radar-based. It lacks the ASW and ASM capabilities Aegis' got. By the way, with a mark 48, that destroyer is hardly ASW compared to M. 54... With Vietnam's recent purchase of 6 kilos, I don't think it can go beyond Xisha in a battle. It is nevertheless a beauty.