Collaborative invention mining at Tata Consultancy Services

Collaborative invention mining at Tata Consultancy Services

Matching inventive ideas to business objectives is a complex interdisciplinary process. However, so-called ‘collaborative invention mining’ can help an organisation to align its inventions with its long-term strategy

When an organisation’s R&D unit comes up with a new inventive idea, multiple stakeholders from the IP, business and product teams must analyse it to decide on its:

  • technical feasibility;
  • usefulness and sustainability to the organisation with regard to its current business worth and future potential;
  • strength and value proposition from the IP portfolio point of view;
  • fit with IP protection strategies;
  • scope against existing prior art to determine novelty; and
  • inventiveness, patentability and possible claims, if patentable.

The idea itself is a result of the inventor’s creative thinking and comes, quite naturally, with little clarity with regard to resulting technology and the engineering required for use in a business application. The journey from idea to commercialisation therefore comprises two steps:

  • designing a solution and making the idea implementable and useable; and
  • analysing the sustainability of the solution developed using the idea.

The challenge lies in finding reliable answers to the above six questions at the idea stage before engaging in these later two activities. In today’s knowledge-driven economy, there is a risk that intellectual property can be lost to other market players if protection for an idea is delayed until these activities are carried out.

This article presents an innovative framework which we have developed at Tata Consultancy Services: collaborative invention mining (CIM). This system maps an inventive idea along certain technical and business dimensions and then asks sets of structured questions from the various perspectives required to achieve an organisation’s goals.

Key benefits

The model-based approach presented here enables the following objectives to be achieved.

Patentability, claim objects and claim construction inputs for inventive idea

CIM provides a detailed view of the three patentability requirements of an invention: novelty (N), inventive step (I) and utility (U). Systematics detailing the invention and documenting findings along these three dimensions (defined and collectively addressed here as NIU) against prior art help to deliver an objective assessment of patentability. This facilitates clear, in-depth visualisation of the technical nature of the inventive idea and the technical solution(s) addressed by the technical problem. It also helps in logically identifying and extracting a complete set of claim objects and their composition when it comes to drafting claims.

Understanding capability and enablement aspects of inventive idea

A focused view of the inventive idea enabled through structured detailing and documenting along the process (P), technology (T), measurement (M) and system (S) dimensions (defined and collectively addressed here as PTMS) helps to provide assurance about technical feasibility and implementation of the inventive idea.

Improving scope of inventive idea

An analysis of NIU and PTMS helps to highlight the gap between initial disclosure and a broader set of possibilities, uncovering hidden claim objects along which the invention’s scope can be expanded.

Sustainability assessment of inventive idea

A structured approach for deliberating and documenting the findings along efficient (E), adaptable (Ad), agile (Ag) and anticipative (An) lines (defined and collectively addressed as EA3) helps when it comes to assessing the sustainability of the business solutions developed using the inventive idea.

Complete view for drafting patent applications

Such detailing of the inventive idea after a prior art search gives the patent agent a complete view when it comes to drafting a patent application.

Invention score and inputs for management decisions

CIM also provides an objective view of the inventive idea in terms of the invention score, which forms the basis for its objective valuation and comparison with other relevant patents. Senior management in large enterprises tends to prefer the technical and business worth of an inventive idea to be presented in numerical terms. A reliable invention score based on the correctly chosen technical and business dimensions obtained through this model-based approach thus provides valuable input for management decisions with regard to portfolio, business alignment, protection and maintenance of the intellectual property.

Ideation tree

The description of a given inventive idea can be broken down into a number of key highlights of the idea derived from the invention description, bearing in mind that they may potentially form components of the claims for the inventive idea. The three tags defined above (ie, NIU, PTMS and EA3) can then be assigned to each of these statements. Such key statements can be placed into one of the 48 cells (3 x 4 x 4) depending upon the NIU, PTMS and EA3 tags assigned to them. For easy visualisation, the 48 categories of invention objects (represented by the 48 cells) may be treated as leaves of the so-called ‘ideation tree’, with its branches classified into three categories: N-PTMS-EA3, I-PTMS-EA3 and U-PTMS-EA3, as shown in Figure 1.

Figure 1. Ideation tree

The success of invention mining lies in identifying as many key statements about the inventive idea as possible and mapping them to the 48 cells. The green bubbles in Figure 2 indicate that cells are filled with at least one tagged key statement, while grey bubbles are empty. A wishful expectation from invention mining would be to fill all 48 cells with one or more key statements, as shown in the tree on the left in Figure 2. However, the tree on the right in Figure 2 may represent a more realistic situation.

Figure 3 shows the transformation of initial key statements into the claim objects for the inventive idea.

Figure 2. Wishful and realistic ideation tree post-invention mining

Figure 3. Transformation of initial key statements to claim objects

A structured process for identifying the key statements, assigning them three tags (one from each of the NIU, PTMS and EA3 sets) and finally mapping them to the 48 cells for computing the invention score is explained in subsequent sections as part of the invention-mining activity.

CIM model

‘Invention mining’ can be defined as the process of maturing an idea to a patentable invention through collaborative interaction. We have defined, implemented and established a collaboration-based invention-mining model to efficiently and predictably functionalise, synergise and measure the invention-mining process.

The CIM model comprises orchestration of the (NIU) x (PTMS) x (EA3) strategy, the 48-cell matrix shown in Figure 4.

Figure 4. Collaborative invention mining (CIM) model

NIU comprises an idea’s patentability criteria: novelty, inventive step and utility.

PTMS comprises an idea’s ‘capability coverage’: process, technology, measurement and system.

EA3 comprises an idea’s ‘sustainability characteristics’: efficiency, anticipability, agility and adaptability, which embed the value in it.

The parameters used for patentability criteria, capability coverage and sustainability characteristics are as follows:

  • ‘Novelty’ refers to an idea’s patentability scope, involving an assessment of the idea to decide whether it is new in the context of pre-existing knowledge in the public domain.
  • ‘Inventive step’ refers to an idea’s patentability, along the line of technical advancement as compared with existing knowledge.
  • ‘Utility’ refers to an idea’s patentability scope, which demonstrates applicability in a new way for a given set of real-world scenarios.
  • ‘Process’ refers to the method or framework deployed by the inventive idea.
  • ‘Technology’ refers to the technical problem addressed by the inventive idea and a technical solution to the technical problem.
  • ‘Measurement’ refers to the special measurements or metrics enabled or deployed by the inventive idea.
  • ‘System’ refers to the implemented system as per the idea’s method or framework.
  • ‘Efficiency’ refers to the ability of the inventive idea enabling the implemented solution(s) to do more with less, making it faster, cheaper or better.
  • ‘Adaptability’ refers to the idea being configurable and flexible for use in several solutions.
  • ‘Agility’ refers to the idea enabling the implemented solution(s) to be able to sense and quickly respond to changing needs.
  • ‘Anticipability’ refers to the inventive idea addressing future problems and thereby resulting in longevity or lasting longer in the market.

CIM process

The CIM process takes the inventive idea through well-structured deliberations which are designed to take account of the various roles which come into play during the journey from idea to commercialisation. These deliberations are carried out in a phased manner. The exercise results in an inventive idea maturing into a patentable invention with enhanced longevity in the enterprise.

The CIM process systematically enhances the capabilities of the inventive idea in an enterprise through structured questions and collaborative throughput aimed at deepening the scope along NIU, broadening the coverage along PTMS and aligning with objectives along EA3 – as depicted in Figure 5.

Figure 5. Collaboration work flow

The CIM working template shown in Table 1 comprises a 48-cell matrix (as per (NIU) x (PTMS) x (EA3) strategy), a parking lot, and notepad-1 and notepad-2, which are used to carry out four CIM phases described below.

Table 1CIM working template

CIM exercise

Before starting a CIM session, the following inputs are required:

  • idea sharing by the inventor(s);
  • initial idea understanding notes; and
  • detailed search results and analysis

The tools used are the CIM working template and the invention scoring template (explained below).

The different roles played by the team include inventor, business manager, IP legal, portfolio manager, patent agent, patent examiner, product manager and moderator.

The CIM exercise is conducted in four well-structured phases: storm, form, norm and compose. Figure 6 shows the structured flow of deliberations during the four phases. Each phase comprises an entry-task-validate-exit sequence. The input conditions, tasks, outputs and validations carried out by the different roles played during each of the four phases are outlined below.

Figure 6. Storm, form, norm and compose – the collaboration phases for CIM

Storm phase

This is first phase of the CIM exercise. Its primary objective is to deepen the scope of the inventive idea along PTMS.

It begins by capturing the invention highlights as key statements in the parking lot area of the CIM working template.

The next step is to identify the T statement, P statements, S statements and M statements from the parking lot entries in the following manner and list them in notepad-1 in the CIM working template.

A T statement simply answers the question: what is the technical problem addressed by the invention? A qualified T statement will typically be made up of the preamble of the independent claims. One inventive idea is expected to address only one technical problem. If we come up with more than one technical problem for an idea, we need to plan separate CIM sessions for each idea. The identified T statement should be entered against T in notepad-1.

P statements for the inventive idea are obtained by focusing on the method (eg, algorithm, steps) for solving the technical problem stated as the T statement. The method should be broken down into multiple P statements (ie, P1, P2, P3), such that each P statement is implementable, with its input, output and interconnects clearly identified. The P statements should then be entered against P1, P2, P3 and so on, in notepad-1.

S statements for the inventive idea are obtained by focusing on system (implementation, components) aspects of the T statement. At this stage, one can safely assume that for every P statement, there will be a corresponding S statement which can be implemented with the requisite engineering and technology inputs. Thus, a given set of P statements will yield a same-sized set of S statements (ie, S1, S2, S3), with one S statement per P statement. The identified S statements should be entered against S1, S2, S3 and so on, in notepad-1.

From a claims perspective, P statements and S statements will typically constitute the portion of the claim suffixed to the preamble. It is crucial that the inventor fully explains the challenges that he or she faced and how he or she solved them for each P statement and S statement – this will help when it comes to identifying inventive step. ‘Obviousness’ should be understood based on whether solutions to the challenges were picked up from prior art, picked up but improved or newly created. Insights into inventive step and obviousness gathered during the CIM session should be noted down separately and incorporated later when drafting the claims and specifications.

M statements for the inventive idea are obtained by finding out the special measurements or metrics enabled or deployed in the P statements and S statements. Useful questions for determining M statements include the following:

  • Do we measure something?
  • What do we measure and is that a novel parameter?
  • What are the key challenges in measuring?
  • How do we measure (ie, using a sensor, transducer, formula or tool)?

The identified M statements should be entered against M1, M2, M3 and so on, in notepad-1. M statements should be used appropriately while drafting the claims and their specifications.

The storm phase concludes with a first review of the text for all statements entered in notepad-1 during this phase.

Form phase

This is the second phase of the CIM exercise. Its primary objective is to broaden coverage of the inventive idea along NIU.

The phase begins by capturing the U statements and entering them in notepad-1 in the CIM working template.

U statements for the inventive idea relate to key industry applications or application categories in which the invention can be used. They are typically used in the specifications section of the patent draft and rarely appear in the claims (unless it becomes necessary to narrow the scope of the claims).

Next, deliberate on the questions about novelty and inventive step in each of the T statement, P statements, S statements and M statements and type ‘Y’ against the respective rows in notepad-2 below the columns labelled ‘N’ (for novelty) and ‘I’ (for inventive step) respectively if the answer is true – otherwise, leave it blank.

Then, deliberate on questions about process, technology, measurement and system tags for all the U statements and type ‘Y’ against the respective rows in notepad-2 below the columns labelled ‘P’ (for process), ‘T’ (for technology), ‘M’ (for measurement) and ‘S’ (for system) respectively if the answer is true – otherwise, leave it blank. The form phase concludes with a review of the entries made in notepad-2 and a second review of the text for all statements appearing in notepad-1 in view of the notepad-2 entries made during this phase.

Norm phase

This is the third phase of the CIM exercise. Its primary objective is to align the inventive idea with the business objectives.

In this phase we question the sustainability of all the statements appearing in notepad-1 in terms of their efficient, adaptable, agile and anticipative characteristics and type ‘Y’ against the respective rows in notepad-2 below the columns labelled ‘E’ (for efficient), ‘Ad’ (for adaptable), ‘Ag’ (for agile) and ‘An’ (for anticipative) if the answer is true – otherwise, leave it blank. The norm phase concludes with a review of these entries in notepad-2 in this phase and a third (final) review of the text for all the statements appearing in notepad-1 in view of the notepad-2 entries made during the norm phase.

Compose phase

This is the fourth and final phase of the CIM exercise. In this phase we map all statements entered in notepad-1 by entering each statement into one of the cells of the 48-cell matrix on the left side of the CIM working template, in accordance with the ‘Y’ entries in notepad-2 against the statement. The number of entries of a particular statement in notepad-1, in the 48-cell matrix, will be equal to the total number of ‘Y’ entries against it in notepad-2. For convenience and usability, only the symbols T and so on, P1, P2, P3 and so on, S1, S2, S3 and so on, M1, M2, M3 and so on, U1, U2, U3 and so on, representing the corresponding statements in notepad-1, need be entered in the 48-cell matrix, rather than the full text of statements.

The compose phase concludes with the composition of a first draft of the invention claims by using the final text statements in notepad-1 and calculating the invention score based on entries in the 48-cell matrix, as per the invention scoring template set out below.

Invention scoring template

The CIM exercise generates a score for the inventive idea. It uses a scheme which takes into account all three leading dimensions (ie, NIU, PTMS and EA3) for calculating the invention score. The invention scoring scheme shown in Table 2 can be used, as explained below.

Table 2CIM scoring template

The scoring scheme calculates the invention score on the basis of the number of statements (appearing in notepad-1 of the CIM working template) mapped into each cell of the 48-cell matrix at the end of the compose phase. It first computes individual scores for each of the 48 cells using the principle of diminishing returns where the cell contains more than one statement it. For example:

  • score for a cell containing one statement = product of the corresponding weights for PTMS and EA3;
  • score for a cell containing two statements = (product of the corresponding weights for PTMS and EA3)*(1+1/4);
  • score for a cell containing two statements = (product of the corresponding weights for PTMS and EA3)*(1+1/4+1/8), and so on.

Once the individual score for all 48 cells is known, three category scores are calculated:

  • N_score = sum of individual cell scores for the 16 cells in the four rows against novelty;
  • I_score = sum of individual cell scores for the 16 cells in the four rows against inventive step;
  • U_score = sum of individual cell scores for the 16 cells in the four rows against utility;
  • aggregate invention score for the idea = computed using the weights for N, I, U respectively and the three categories as shown below; and
  • aggregate invention score (out of 1,000) = [4*N_Score + 3*I_Score + 3*UScore].

CIM system

Figure 7 presents the system view of the implemented CIM model. The CIM system rests on a collaboration framework.

Figure 7. CIM workbench

It consists of an input interface for receiving the inputs from all stakeholders when they perform their key tasks across all stages of the collaboration. There is an idea management module which manages and tracks the ideation data received from the inventors. There is a search and analytics module for various types of search and analytics needs. The ideation tree is the core of the invention-mining system for bringing dimensional rigour to the idea and turning it into a patentable invention.

The claim construction module (which is not elaborated here) is designed to help compose the invention claims with the help of claim objects obtained as output of the CIM exercise. There is a metrics generation module for generating the invention score and related analysis. It also consists of the display and visualisation module for different views and reporting needs. The system is integrated with the organisation’s IP management system.

Action plan

Designing and developing a good invention-mining system is necessary for any IP-based organisation in order to ensure that its inventions are technically deployable and commercially sustainable:

  • First, build the organisation’s invention-mining capability to define a mandate for invention mining as part of its patenting work flow.
  • Design the invention-mining framework by customising the CIM model in order to bring dimensional rigour to the process of maturing the idea into a patentable invention, according to the nature of the organisation’s business and its priorities.
  • Conduct manual CIM exercises for the organisation’s inventions to understand the needs for further customisation of the model and derive business requirements for its automation.
  • Develop and implement an automated system for invention mining in the organisation.

Santosh K Mohanty is vice president and Akhilesh C Srivastava leads IP deployment and support for corporate R&D at Tata Consultancy Services, Mumbai, India

Unlock unlimited access to all IAM content