STUDY ON MAKING TECHNICAL ROADMAP ATTRIBUTED TO THE INFORMATION OF PATENT
Jina Lee¹,Ji-Youn Lim², Ji-Young Lee², Sang-Chan Park³
(1),(2),(3)Department of Industrial & Systems Engineering,
Korea Advanced Institute of Science and Technology (KAIST),
373-1 Guseong-dong, Yuseong-gu, Daejeon, 305-701, Republic of Korea
ABSTRACT
Increasing uncertain information of environment, it is going harder to make a priority and coordination of technologies. To make a decision, Technology Road Map (TRM) has been. TRM is an instrument that supports strategic technology management and planning. TRM provides a mechanism and information to help make better technology investment decision. Initial step of TRM is making priorities of technologies. And this step is generally based on the rating of core technology that is evaluated by experts committee. TRM is designed with these priorities. This method is useful and commonly used, but there are limits, it could be biased by each expert’s characteristics. To overcome this problem, we suggest the road-mapping process, as a computer based approach, with using the information of patent. In this method, the number of patentsis used to value the importance of each technology. The advantage of this method is expected to take the scores by objective data and excluding private opinion of expert. This paper presents the method of evaluating the scores of each technology as initial step of building up TRM.
Keywords: Technology Roadmap (TRM), Computer Based Approach,Patent
INTRODUCTION
The role of services has been a long term trend over many decades. The human history has been developed from hunting society to service industries. The steps of shift are as follows: hunting society → agriculture society until 17th century → manufacturing industry →information-oriented society→ service industry. In hunting society and agriculture society, human life was mainly dominated by materials. After the industrial revolution, materials and energy has been staying important positions in human life. However, after information revolution occurred in the 1950s, the information-oriented society has been started. Information-oriented society is the society that is dominated by not only material and energy but also information. Especially information has strong power.In information-oriented society, abstract ‘Knowledge’ or ‘Information’ takes the core part of manufacturing and distribution, more than manufacturing and distribution of materials. And the value of information increases steadily. According to information society, service economy has been rapidly growing both qualitatively and quantitatively in the global economy.(Chesbrough & Spohrer, 2006). The importance of service economy is evident as follows.
Figure 1. Service industry in South Korea
Table 1. Top 10 countries by workforce size in 2005 (source : IBM)
Country / % WWF / % Agriculture / % Manufacturing / % Service / % 25 year growthChina / 20 / 50 / 15 / 35 / 191
India / 17 / 60 / 17 / 23 / 28
US / 4.8 / 3 / 27 / 80 / 21
Indonesia / 3.9 / 45 / 16 / 39 / 35
Brazil / 3 / 23 / 24 / 53 / 20
Russia / 205 / 12 / 23 / 65 / 38
Japan / 2.4 / 5 / 25 / 70 / 40
Nigeria / 2.2 / 70 / 10 / 20 / 30
Bangladesh / 2.2 / 63 / 11 / 26 / 30
Germany / 1.4 / 3 / 33 / 64 / 44
The service economy has grown significantly in both developed and developing countries though the ways of their shifting to the age of services are different (Paulson,2006). In case of China, the growth rate of service is 191 percent during last 25 years. Most of countries, US, Germany, Brazil, Japan, make up more than 50 percent of the whole economies as Table 1. Also in South.Korea, service accounts over 50 percent of economy after 2000.
As service industry becomes bigger, paradigm has shifted from products and materials to invisible sources and services(Vargo & Lusch, 2004). Regarding this change, the role of R&D is becoming more important.
There are three R&D areas in services, service R&D, education and strategy & policy (Paulson, 2006). First one is service R&D. Service R&D is also called as service science, and it is central field in industry R&D. Due to the importance of service, the development in service R&D area causes innovation and growth of production. Service R&D is a brand new concept, many universities and companies are studying for clear definition. Second, in education field, many universities in Europe, North America and Australia, provide curriculums and certifications for development service science. Not only universities, but also companies, for example IBM,recommend the education in many fields as technology management, management science and engineering. The last area is strategy and policy fields. The purpose of strategy and policy is supporting the service companies for efficient management by choosing the proper technology and service, and by concentrating them. However, method of analysis tool to achieve such purposes does not exist correctly in service industry. Technology Roadmap can show the guidelines for choice and concentration. With this reason, technology roadmap is needed as method and analysis tool for R&D strategy. Nowadays, experts give points to each technology, and importance of each technology is determined by these points.Making priorities among technology factors is important at the initial step of roadmapping. Based on the priorities, the expert groups design the technology roadmap. (Groenveld, 1997). With this method, the priority of technology can be changed greatly by the private opinions. This paper suggests the methodology that can determine the importance of each technology using patent information.
LITERATURE REVIEW
Definition of Technology RoadMap
Science and technology roadmaps are used to portray the structural relationships among science, technology, and applications. Roadmaps are employed as decision aids to improve coordination of activities and resources in increasingly complex and uncertain environments(Ronald N.Kostoff and Robert R.Schaller, 2001). Roadmap is an artifact that reflects a common vision in a particular field and for a desired objective. Technology Roadmap is a useful instrument that supports strategic technology management and planning, and provides a framework for supporting integrated and aligned multifunctional strategic planning(Roberta Cuel, 2005).Technology roadmapping is not an isolated independent approach, but rather a method that supports strategic planning at the firm, sector or national levels ( Robert Phall et al., 2004). Technology roadmapping is different from strategic planning. Roadmapping is seen as a discrete step in the strategic planning process as a key step towards implementation. Strategic planning is a mature process, roadmapping concepts tend to ‘infiltrate’ the whole strategic planning process.One of the key benefits of roadmaps is to provide a focus for integration within strategic planning and innovation processes.
Many roadmap variants can be aggregated into two fundamental roadmapping approaches : expert-based and computer-based.Expert-based approach is to draw on the knowledge and experience of the participants to identify the structural relationships within the network and specify the quantitative and qualitative attributes of the links and nodes. Computer-based approach generates the network at all points in the time domain of the source database simultaneously. So it has more objectivity than the expert-based but it is in its infancy now.Kappel(2001) classified technology roadmaps into four types by the main subject of making roadmap (nation, company) and target analysis (science and technology, product). Roadmaps can be classified classify the roadmaps by time domain as dynamic and static roadmaps( Albright and Kappel, 2003).
Patent information
Patents are used for measuring technologies changes. Like other indicators, patents have both advantages and disadvantages. Patents show directly about inventive process, especially those inventions which are expected to have a commercial impact. They are a particularly appropriate indicator for capturing the proprietary and competitive dimension of technological change. Patents are broken down by technical fields and thus provide information not only on the rate of inventive activity, but also on its direction. Patents are public documents, so all information, including patentees’ names, is not covered by statistical confidentiality. With these advantages, we use patent information for making technology roadmap.
FRAMEWORK & METHODOLGY
There are a lot of analysis tools for present technology, but method for emerging technology is not sufficient.It is not easy to estimate the services for emerging technology unlike estimation of present technologies for the products and materials. Information of patents are offering the trend of technology, so patent analysis is proper for technology roadmap. It is obvious that knowing the technology trend is important for reducing the unnecessary investments and developing the adequate technology at the right time. To achieve the purpose, we suggest the method for rating each technology in five indexes as marketability, growth-ability, imminence, origin-ability and affectability. Complicated network is needed for containing all of the patent information. And the network is difficult to understand and hard to apply directly.Avoiding this problem, we suggest the table of rating information not focusing on the visualizing presentation. This method has advantages. It is more objective method as using patents for rating technologies. And it can be applied in system immediately.
Figure 2. Flow chart of patent analysis
For using patent information for grading the technologies, keywords of each technology should be determined. Each technology can have several keywords. With these keywords, we can rate the technologies as searching the patents each year.
Origin-ability
Origin-ability means how much the technology is powerful in being the source of other technologies. We define the origin-ability is the number of patents which have exactly one keyword of the technology. With this origin-ability index, we can judge the nature of each keyword by the number of patents including them. As Figure 3, the origin-ability of the technology is the sum of each keyword’s patents. It is the proportion of gray part from the whole.
Figure 3. Venn diagram for Origin-ability
If a technology has 3 keywords, for example [a], [b], and [c], then origin-ability is gained as sum of each keyword’s origin-ability index. The index of keyword [a] can be obtained as searching patents including [a] and excluding [b] and [c]. Just like this step, we can acquire the origin-ability index.
For example, the origin-ability of “Defect-free casting technology of low thermal expansion bed for wafer inspection equipment” is showed as follows.
Keywords
:defect free, casting, low thermal expansion, bed, wafer, inspection
Number of patents containing each keyword and excluding other keywords
:defect free = 100
:casting = 4,977
:low thermal expansion = 45
:bed = 2,770
:wafer = 27,558
:inspection = 20,435
Origin-ability = 100 + 4,977 + 45 + 2,770 + 27,558 + 20,435
= 55,885
We consider the technology has an origin power if index is larger than the average of whole technologies in the same field. The priority among technologies in same field can be obtained comparing the indexes. And we are also able to decide the priority among other field technologies with the normalized index by dividing the average origin-ability index of each field.
Growth-ability
Growth-ability represents the growth of technology literally. This index is determined by the number of patents from the past to current time and increasing rate of patents. Number of patents is not constant. It varies by the trend of the time and technological levels. It tends to flow out heavily in short time and decrease in the course of time.
Based on this tendency, we should determine three measures, weight, threshold and link proportion.First, different weight is assigned to each year. The recent past is considered more important than the remote past. By granting different weights, we can estimate the level of growth and the ability of growth in the future. Second,we have to set the threshold for eliminatinganegligiblenumber of patents. We assume that the year exceeding threshold is meaningful with the keyword. And last, the link proportion is determined by the existence of patents which have combinational keywords. The equation of growth-ability is an weighted average including all of these measures. In this paper, we consider the period of growth for 15 years, from 1992 to 2007, as long-term tendency.
(1)
: number of keywords ini technology
: growth rate of j keyword in i technology
: weight of j keyword in i technology
(2)
: the first year of period, 1992
: the last year of period, 2007
: instant increasing rate of j keyword in i technology at time t, (t = 0,1,….,12)
The growth rate of keyword can be obtained as the sum of each year’s instant increasing rate, equation (2). And time variable, t, is multipliedto instant increasing rate due to granting more importance of the recent past. The time variable of year 2007 is 15, and year 1992 is 1. The equation of instant increasing rate is as follows:
(3)
:number of patents, j keyword in i technology at year u, (u = 1992,…,2007)
Figure 4. Graph of instant increasing rate
We define the weight as multiplying three values, link proportion, and frequency more than threshold and first year that related patent had occurred. With these values, weight gets the meaning of time tendency, as mentioned above.
(4)
: link proportion of j keyword in i technology
: frequency that exceeds the threshold, j keyword in i technology
: time variable t for the first year that exceeds the threshold,j keyword in i technology
(5)
: number of linking keywords related j keyword in i technology
: maximum of
Link proportion is the number of linking keywords, divided by the maximum number of link keywords in each technology. According to dividing maximum number, we can get the normalized link proportion through the whole technologies. For example, “Defect-free casting technology of low thermal expansion bed for wafer inspection equipment”has 6 keywords. We can find the patents including 3 keywords waferinspectiondefect free. But there is no patent during 15 years containing more than3 keywords. There are patents with 2 keywords, low thermal expansion & wafer, or bed & inspection, or casting & wafer. The three keywords, wafer, inspection and defect free, have 3 linking keywords, including itself. The other three keywords, low thermal expansion, bed and casting, have 2 linking keywords of patent.So the maximum number is 3 in this technology. In this case, the linking proportion is 3/3 for the first three keywords, 2/3 for the last three keywords.
(6)
: 1 when number of patents exceeds the threshold , j keyword,i technology, at time t
0 otherwise
The frequency is the sum of years that exceeds the threshold. We assume that threshold is a minimum number as a proof that keyword was meaningful or negligible. The year, related patents occurred less than threshold is negligible. And the first year that number of patents exceeded the threshold is . For granting more weights to the recent past, we give the value of time variable into .
For understanding these meanings, detail explanation is followed with an example.
= 6 ( defect free, casting, low thermal expansion, bed, wafer, inspection )
of each keyword is showed as following table and we marked the , whether number of patents exceeding the threshold or not, in italics.
Table 2. Instant increasing rate of each keyword,
year / t / inspection / defect free / bed / wafer / low thermalexpansion / casting
1992 / 0 / 433 / 1 / 89 / 327 / 1 / 0
1993 / 1 / 456 / 8 / 73 / 575 / 4 / 5
1994 / 2 / 654 / 5 / 84 / 686 / 2 / 99
1995 / 3 / 1189 / 9 / 131 / 1412 / 1 / 269
1996 / 4 / 1421 / 3 / 138 / 2079 / 2 / 405
1997 / 5 / 1773 / 5 / 198 / 1904 / 2 / 360
1998 / 6 / 1420 / 5 / 161 / 2083 / 3 / 377
1999 / 7 / 1167 / 6 / 181 / 2091 / 2 / 444
2000 / 8 / 1358 / 5 / 197 / 1888 / 5 / 502
2001 / 9 / 1470 / 9 / 192 / 2330 / 2 / 478
2002 / 10 / 1476 / 10 / 233 / 2050 / 2 / 484
2003 / 11 / 1734 / 8 / 217 / 2572 / 4 / 451
2004 / 12 / 1878 / 5 / 241 / 2684 / 4 / 389
2005 / 13 / 1922 / 11 / 220 / 2592 / 2 / 302
2006 / 14 / 1527 / 5 / 218 / 1542 / 6 / 265
2007 / 15 / 557 / 5 / 197 / 743 / 3 / 147
The other indexes are showed in Table 3, and GA₁is the growth-ability of each keyword. Growth-ability for the technology, GA₂is sum of the GA₁divided by .
Table 3. Growth-ability of technology
inspection / defect free / bed / wafer / low thermalexpansion / casting
/ 171752.0 / 811.0 / 24189.0 / 234657.0 / 386.0 / 42428.0
/ 1.0 / 1.0 / 0.7 / 1.0 / 0.7 / 0.7
/ 15.0 / 0.0 / 15.0 / 15.0 / 0.0 / 14.0
/ 1.0 / 0.0 / 1.0 / 1.0 / 0.0 / 2.0
/ 15.0 / 0.0 / 10.0 / 15.0 / 0.0 / 18.7
GA₁ / 2576280.0 / 0.0 / 241890.0 / 3519855.0 / 0.0 / 791989.3
GA₂ / = 1188335.7
Imminence
Imminence index is defined that the urgency of the technology, range from 0 to 1.It can be calculated with the number of combinational keywords, as linking proportion. If there is a registered patent including all the keywords, the technology is sufficiently developed. For example, technology 1 has 4 keywords and there are patents with 3 keywords, but no patent including 4 keywords. Technology 2 has 5 keywords and there is no patent including more than 2 keywords. So the maximum number of combinational keywords is 2. With this information, we can calculate the imminence of each technology. Technology 1 is 0.25, divide (4-3) by 4, and technology 2 is 0.6, divide (5-2) by 5. The equation of imminence index is followed :
(7)
The imminence for “Defect-free casting technology of low thermal expansion bed for wafer inspection equipment” is 0.5, divide (6 – 3) by 6.
Marketability
We assume the level of ongoing researches for the technology and the level of technology in market with marketability index. It is the sum of patent number including more than 50 percent keywords. If there are 4 keywords, we count the number of patents having more than 2 keywords. With “Defect-free casting technology of low thermal expansion bed for wafer inspection equipment”, there are only 2 patents, year 1997 and 1998, including more than 3 keywords. So, the imminence for the technology is grading as 2.
Affectability
We can know the availability for the technology whether it can be applied in various fields in forward with affectability index. We collect the patents having more than 50 percent keywords and count the number of extra keywords in the name of patents, using computer programs. The affectability for “Defect-free casting technology of low thermal expansion bed for wafer inspection equipment” is 7 with extra keywords named semiconductor, chip, inferiority, yield loss, pattern, plate and measurement.
CONCLUSION
We analyzed the 139 technologies in production-based field. There are 765 technologies in this field, but due to the purpose for analysis is to find the important technology, it is proper to limited in 139 level A technologies. As Table 4, we can see the max, min, average and standard deviations of five indexes.
Table 4.Summary for analysis result
Marketability / Affectability / Origin-ability / Growth-ability / ImminenceMaximum / 31308 / 119728 / 359947 / 20143395 / 1
Minimum / 0 / 0 / 42 / 0 / 0
Average / 969.87 / 4364.13 / 69358.09 / 2542420.50 / 0.40
Standard Deviation / 3255.49 / 15411.92 / 78701.25 / 3713017.33 / 0.19
Conversion average / 3.16 / 2.17 / 18.60 / 12.78 / 39.62
For normalizing five indexes, we used to converse the average based on 100 scores. The affectability and growth-ability record about 10 points, marketability and imminence are remarkably low, 3.16 point and 2.17 point. We estimate that most of technologies have relatively low scores and a few technologies have extremely high scores. Different from the other indexes, high imminence score does not mean that the technology has been developed a lot. It means the technology should be developed in short time.