WF-20 — Intellectual Property Disclosure and Protection

KPI 4.6 counts granted IP assets, and copyright covering educational works is explicitly inside the definition. That is the fact that makes this workflow interesting for an institute of learning rather than a formality borrowed from a science faculty. IoL generates software, assessment instruments, simulation scenarios, curricula and educational materials continuously, and almost none of it is ever put through a disclosure process, because nobody thinks of a course as an asset. Ten granted assets scores 100 and the anchors start at one, so a single granted asset moves a KPI that is plausibly reporting zero today. The honest counterweight, stated at the top so it is not buried: registration costs money and takes time, and the decision to spend either belongs to technology transfer and legal advice, not to a reporting process. This workflow's job is to make sure the decision is taken, on the right things, by the right people. It is not to file for the sake of a count.

1. Purpose

To ensure that intellectual property created by IoL staff and learners is identified and disclosed when it is created, assessed on its merits for protection by the people qualified to make that assessment, and where protected, recorded accurately as a granted asset with MBRU's ownership share, its jurisdiction and its grant year, so that the institution knows what it owns, the creators are treated properly, and the count MBRU reports is one it can evidence.

2. Scope

Scope statement. This process manages intellectual property from the point at which a potentially protectable output is created or is about to be disclosed publicly, through disclosure, ownership determination, protection assessment, filing and prosecution, to the recording of the grant, its maintenance and its reporting.

Applies to. All IP created by IoL academic and professional staff in the course of their duties, and by PGDipHPE and MScHPE learners where MBRU has a share under the applicable policy [IoL to confirm the MBRU IP policy and its treatment of student-created IP]. All eligible categories: registered patents, plant variety protection rights, granted utility models, integrated-circuit layout designs, industrial designs, and copyrights covering software, research-based works, educational works and creative works. Specifically including educational outputs: e-learning software and tools, assessment instruments and item banks, simulation scenarios and their supporting materials, curricula and structured educational programmes.

Does not apply to. Publication of research findings, which is WF-18; articles in peer-reviewed journals are expressly excluded from the KPI 4.6 copyright category and must never be counted here. Research impact as an effect in the world, which is WF-19, although a granted trademark or a concluded licence properly evidences a KPI 4.5 commercialisation unit as well as being an asset here. Revenue from licensing as an industry contribution total, which is WF-24 and KPI 3.4. Trade secrets and confidential know-how, which are managed but are not granted assets and do not count.

Applicable requirements. OBEF guide v11.5 KPI 4.6; CHEDS Institute - Intellectual Property.xlsx and the HEDB lookups IP Type, Patent Status and Patent Office; CHEDS Master API specification 18 February 2026, fields OBF4.6.N, OBF4.6.D, OBF4.6.R, OBF4.6M; UAE intellectual property law and the requirements of each registering office in which MBRU files; MBRU IP and commercialisation policy [IoL to confirm]; contractual IP terms in research funding and partnership agreements.

The definitions that govern everything below.

What counts. Registered patents; plant variety protection rights; granted utility models; integrated-circuit layout designs; industrial designs; and copyrights covering software, research-based works, educational works and creative works.

What does not count. Applications still pending in any country. IP owned solely by individuals at the institution. And within the copyright category, articles published in peer-reviewed journals and lecture notes, which are explicitly excluded.

The ownership test. MBRU must hold at least a 20% share of the IP. This is a contractual question settled long before the grant, in employment terms, funding agreements and partnership contracts, which is why section 8 treats it as a design action rather than a reporting check.

The counting rules. IP counts only in the year it was granted, and cannot be re-counted in a later year. Where the same IP is granted in more than one jurisdiction, each instance counts separately. That second rule is the one place in the framework where apparent duplication is correct, and it cuts both ways: an extract that de-duplicates on title will understate the count, and a reviewer who has not read the rule will read the correct extract as double counting. Say so in the mechanism field.

A genuine ambiguity in the guide, flagged rather than resolved. The KPI counts granted assets and the CHEDS template carries Status, Approval Date, IP Number and Registering Office, all of which presuppose a registration act. Copyright, however, arises automatically on creation in most jurisdictions and does not require registration to subsist. The guide does not say whether an unregistered copyright can be a "granted IP asset" for KPI 4.6, and the template offers no way to describe one. The conservative reading, which this workflow adopts until MoHESR says otherwise, is that a copyright counts only where it has been registered with a registering office that issues a dated record, because that is the only reading the template can carry. Put the question to MoHESR in writing before relying on either reading, because the answer changes the size of the educational IP opportunity in section 8 substantially.

3. Trigger, boundary and endpoint

4. SIPOC

5. Accountability

Process owner. MBRU technology transfer or research commercialisation lead, with the IoL Research Lead as the departmental counterpart. This is a deliberate split. IoL owns the disclosure discipline and the identification of educational outputs worth assessing. It does not own the decision to file, the choice of jurisdiction, or the legal assessment, and it must not, because those decisions commit money and carry legal consequence.

Escalation.

6. Process steps

1. Prompt disclosure at the moment of creation, not at the moment of commercial interest. Attach a standing disclosure prompt to the points where IoL actually creates protectable material: release of a piece of educational software or a tool; completion of a new assessment instrument or item bank; completion of a simulation scenario package; approval of a new curriculum or structured programme; creation of a substantial educational resource intended for use beyond MBRU. [CAPTURE] the prompt response, including a negative one. Serves 4.6. Decision point. If public disclosure (a conference presentation, a publication, a public deployment) is imminent and the output may be patentable, the disclosure must be made before the public act, and the public act held until triage. Novelty destroyed is not recoverable. For copyright and for most educational outputs this urgency does not apply, which is worth saying so the rule is not applied indiscriminately and does not become a brake on teaching.

2. Submit the disclosure. Creator, date of creation, description, contributors and their contributions, funding sources and any agreement that may govern ownership, whether it has been publicly disclosed and when, and any third-party material embedded in it. [CAPTURE] disclosure record with a unique reference and a receipt date. Serves 4.6 and the ownership test.

3. Log the disclosure and start the triage clock. Every disclosure receives a dated decision, including a decision not to protect. A disclosure that receives no decision is the failure mode that makes creators stop disclosing. [CAPTURE] log entry, triage owner, target decision date. Serves 4.6.

4. Determine ownership and MBRU's percentage share. Read the employment terms, the funding agreement and any partnership contract. Record the share MBRU holds and the shares held by others. [CAPTURE] Applicants/Owners, Percentage share of the HEI in the IP, Inventor IDs, Inventor ORIC IDs. Serves 4.6. Two hard tests, both applied here rather than at extract. MBRU must hold at least 20% for the asset to count. And an asset owned solely by individuals at the institution does not count at all, however meritorious. Both are contractual outcomes, so the place to fix them is the agreement, not the register.

5. Assess protectability, cost and value. Technology transfer, with legal advice, assesses novelty and protectability, the categories available, the jurisdictions worth filing in, the cost of filing, prosecution and renewal, and the commercial or strategic value. [CAPTURE] the assessment, its date and its author. Serves 4.6. State plainly what this step is not. It is not a scoring calculation. A weak asset filed to raise a count consumes money and attention for as long as it is maintained, and the framework does not reward it beyond the single unit it contributes. The assessment must be capable of returning "do not protect" without reference to the KPI, and it must do so often.

6. Decide, and record the decision either way. Protect, protect later, or do not protect, each with a reason and a date. Where the decision is not to protect, record whether the output should instead be released openly, held as confidential know-how, or simply used internally. [CAPTURE] decision, reason, date, decision maker. Serves 4.6 and the audit trail. Exception route. Where MBRU declines to protect, the applicable policy may permit assignment to the creator [IoL to confirm]. If that happens, note that the resulting asset is then owned solely by an individual and does not count for KPI 4.6, which is a correct outcome and not a loss to be avoided.

7. File and prosecute. Managed by technology transfer with external counsel. Record the application number, the filing date, the registering office and the designated country, and hold Status at pending. [CAPTURE] IP Application Number, filing date, Registering Office, Application Designation Country, Status = pending. Serves 4.6 pipeline visibility only. A pending application is not a counted asset, in any country. Keeping pending rows in the register is correct and useful; counting them is not, and the extract filter in step 12 is what prevents it.

8. Record the grant, which is the only event the KPI recognises. On grant, record the approval date, the IP number, the expiration date and set Status to granted. [CAPTURE] Approval Date, IP Number, Expiration date, Status = granted, Year derived from the approval date. Serves 4.6, and this is the capture that produces the number. Approval Date drives the counting year, not the filing date. An asset filed in 2022 and granted in 2026 counts in 2026 and only in 2026.

9. Record each further jurisdiction as a separate instance. Where the same IP is granted in a second or third jurisdiction, create a separate register row sharing the title but carrying its own Application Designation Country, Registering Office, IP Number and Approval Date. [CAPTURE] the additional instance. Serves 4.6. Do not de-duplicate on title at extract. Rows with the same Title and different Application Designation Country are retained as separate counts. Note this explicitly in OBF4.6M, so that a reviewer reading the extract does not mistake correct practice for double counting.

10. Maintain the asset. Track renewal deadlines, pay or decide not to pay, and record lapses and expiries. [CAPTURE] renewal dates, renewal certificates or applications, lapse date and reason. Serves 4.6 and the Appendix B evidence set, which names renewal certificates explicitly. A lapsed asset was still granted in its grant year and still counts for that year within the window. Lapsing does not retrospectively remove a count, and the register must be able to show that.

11. Apply the copyright exclusions before anything reaches the extract. Every copyright row is checked against two exclusions: no articles published in peer-reviewed journals, and no lecture notes. [CAPTURE] the exclusion check, its date and who performed it. Serves 4.6. The boundary between an excluded set of lecture notes and an includable educational work is a judgement, and it should be made once, written down, and applied consistently. A reasonable line, offered as a working position rather than as a rule from the guide: material prepared for the delivery of a specific taught session is lecture notes; a structured, versioned educational product designed for use beyond the session in which it originated, and registrable as such, is an educational work. Record the line taken in OBF4.6M and hold it constant. Getting this wrong in the permissive direction invites a finding on a KPI where the count is small enough that one bad row is a large proportion.

12. Extract, reconcile and release to WF-26. Filter Status to granted only; test Percentage share of the HEI in the IP at or above 20 on every row; confirm Approval Date falls inside the five-year window; retain multi-jurisdiction rows; confirm Applicants/Owners includes MBRU as an institution and not only named inventors; and reconcile with WF-19 and WF-24 so that a trademark or a licence is described consistently wherever it appears. [CAPTURE] extract date, filters applied, row-level exceptions. Serves 4.6, 4.5 and 3.4. The same ambiguity noted in WF-19 applies here. KPI 4.6 is a count, not a ratio, yet the submission carries OBF4.6.N, OBF4.6.D and OBF4.6.R. What belongs in the denominator is not stated. Ask CHEDS, record the answer in OBF4.6M, and apply it consistently across years.

7. OBEF data generated

Capture rule. Three things must be recorded at the time and cannot be reconstructed. The disclosure date relative to any public disclosure, because it is the fact that determines whether patent protection was ever available and it is the first thing an attorney asks. The grant date and the registering office, because they fix the counting year and the jurisdiction, and a register that holds a title and a vague year cannot be submitted. The ownership share as agreed, because it is settled by contract before the grant and reconstructing it afterwards means re-reading agreements that may no longer be findable.

Reproducibility test. A second analyst can reproduce KPI 4.6 from Institute - Intellectual Property.xlsx alone: filter Status to granted, test Percentage share of the HEI in the IP at or above 20, filter Approval Date into the five-year window, count rows without de-duplicating on title. That is a clean test and this KPI passes it provided the register exists. On the assumption that MBRU's Pillar 4 gap is 4.5 or 4.6, the question is not whether the KPI is reproducible but whether there is anything to reproduce it from [IoL to confirm whether an institutional IP register is currently maintained and populated].

8. Max-score design

The IoL argument, made carefully.

An institute of learning generates a category of intellectual property that a science faculty does not: educational IP. Software and e-learning tools. Assessment instruments and validated item banks. Simulation scenarios and their supporting packages. Curricula and structured educational programmes. Educational resources designed for use beyond the institution. This is a real category of output, it is produced continuously, and it is almost never disclosed, because a course does not feel like an asset and nobody has ever asked the person who wrote it to fill in a form.

Educational copyright is explicitly within the KPI 4.6 scope. The guide lists copyrights covering software, research-based works, educational works and creative works among the counted categories, and excludes only journal articles and lecture notes from that category. A systematic disclosure process for educational outputs is therefore a plausible route to points that IoL is unusually well placed to take, and it is a route that a medical or science faculty inside the same university mostly cannot take, because they do not produce this kind of output at IoL's rate.

Four things must be said alongside that argument, or it becomes irresponsible.

1. Registration costs money and takes time. Every filing carries professional fees, official fees and a renewal tail that runs for years. A count of ten assets is also a maintenance obligation.
2. The decision belongs to technology transfer and legal advice, not to this workflow and not to the OBEF pipeline. IoL's role is to identify candidates and disclose them. Whether any of them is protected is somebody else's professional judgement, and the answer will frequently and correctly be no.
3. The copyright registration ambiguity in section 2 is unresolved. If MoHESR confirms that only registered copyright counts, the educational IP route depends on the availability, cost and practicality of a copyright registration route for each output, which must be established with legal advice before any expectation is set. If the Ministry takes the broader reading, the opportunity is considerably larger. Do not build a plan on either answer until it has been given.
4. The point is worth 2.0% of the institutional score and nothing at programme level. That is a real but modest prize, and it should not be allowed to distort IoL's educational priorities. Disclose because knowing what the institution owns is good practice; protect where protection is genuinely warranted.

What this workflow must do to reach the target.

1. Create the IP register with the CHEDS column set as its schema, so the extract is a filter over the register rather than a construction exercise.
2. Attach disclosure prompts to the five points where IoL creates protectable educational material, and accept negative responses, so that the prompt is cheap and therefore survives.
3. Give every disclosure a dated decision within 20 working days, including the decision not to protect. Nothing kills a disclosure process faster than silence.
4. Fix the 20% ownership share in the agreements, not in the register. Every research funding agreement, partnership contract and staff arrangement that could produce IP should be read for the share before signature, because after the grant it is too late.
5. Record each jurisdiction separately, and explain the rule in the mechanism field, so that correct practice is not read as double counting.
6. Apply the two copyright exclusions before the extract, with the lecture-notes line written down and held constant.
7. Reconcile with WF-19 and WF-24 before release, so a trademark, a licence and its revenue are described consistently in all three places.

[REDESIGN] actions.

Sequencing note. R5 and R6 come first in importance even though they produce no points, because they determine whether R2 is a modest administrative addition or a funded programme. R1 and R7 are worth doing regardless of both answers, because an institution that cannot list what it owns has a problem that is not about scoring. R3 is the one with the longest tail: it only affects agreements signed after it is adopted, so every month of delay is a set of agreements that cannot be fixed later.

9. Indirect strategy where data cannot be collected

Category 3 and Category 4 both apply, and distinguishing them correctly is the whole of the analysis.

Category 3, the activity genuinely does not exist. For patents, plant variety rights, utility models, integrated-circuit layouts and industrial designs, the honest position for an institute of learning is probably that IoL does not generate them. A department that teaches and researches health professions education is not going to file a utility model. Say so, and do not manufacture the activity. There is no weight redistribution available on KPI 4.6 (the guide provides redistribution in exactly four programme-level cases, and KPI 4.6 has no programme level at all), so an absent category is simply an unearned count. Nothing is gained by claiming otherwise and there is no classification trap here, which is worth stating because there is one at KPI 5.2.

Category 4, the activity exists but is invisible. This is the educational IP case and it is where the real content of this workflow sits. IoL creates software, assessment instruments, simulation scenarios, curricula and educational resources. Some subset of that is protectable. None of it is currently disclosed, so no assessment is ever made, so no decision is ever taken, so nothing is ever granted, so the count is zero and the reason is procedural rather than substantive.

The response is a disclosure process, and it stops there. Note the difference from every other Category 4 case in this pilot. Elsewhere, capture is sufficient: register the placement, register the event, register the student project, and the KPI numerator moves. Here capture is necessary but not sufficient, because between the disclosure and the counted asset sit a professional assessment, a spending decision and a registering office. This workflow can guarantee that the decision gets taken. It cannot guarantee, and must not promise, that anything gets granted. Any business case built on this workflow should say that in those terms.

What IoL should and should not conclude from that.

• Should: disclose systematically, because an institution ought to know what it has created, and because a candidate that is never disclosed is never assessed.
• Should: get the answers to R5 and R6, because the size of the opportunity is unknown until they are answered and any estimate offered before then is invented.
• Should: fix the ownership share in agreements now, because it cannot be fixed later.
• Should not: file weak assets to raise a count. Two percentage points of institutional weight does not justify a maintenance obligation on something nobody will use.
• Should not: present educational IP as a solved route in a committee paper. It is a plausible route, contingent on a legal answer and a Ministry answer, and it should be described exactly that way.

Category 1 does not apply. MBRU submits KPI 4.6 itself through the Master API. There is no Ministry-held instrument and no third-party database standing between IoL and the number. Whatever the count is, it is MBRU's own record that produces it, which means every failure on this KPI is an internal one.

Category 2 does not apply. The KPI is a count of assets, not a survey of a population, so there is no sampling threshold and no Appendix C grouping question.

Category 5, performance genuinely low, is the likely honest answer for the first year or two. A register created in year one will show a small number, possibly zero. That is an accurate statement of the position and it should be submitted as such, with a corrective action plan through WF-04 if the institution decides the position should change. A zero that is reported is worth more than a zero that is left null, because the operating assumption in this pilot is that an unsubmitted KPI scores zero anyway, and a reported zero demonstrates a working process while a null demonstrates nothing.

Boundary check. This workflow must never:

• count an application that is still pending, in any country, however certain the grant appears;
• count an asset owned solely by individuals at the institution, including one assigned to a creator after MBRU declined to protect it;
• count an asset in which MBRU's share is below 20%, and never round a 15% share up;
• count an article published in a peer-reviewed journal under the copyright category, which is expressly excluded, and never move a publication here because KPI 4.1 did not capture it;
• count lecture notes as an educational work, or blur the line between them after seeing the count;
• de-duplicate multi-jurisdiction grants on title, since each instance counts separately, nor create jurisdiction rows for filings that were never made;
• re-count an asset in a later year because it remained in force, since IP counts only in its grant year;
• date a count from the filing date rather than the Approval Date;
• count an asset whose grant date falls outside the five-year window;
• count trade secrets, confidential know-how or unregistered material where the conservative reading of the guide is being applied, without first recording that the reading has changed and why;
• file an application principally in order to raise a KPI count, or allow the KPI to appear anywhere in the protectability assessment at step 5;
• record a disclosure decision retrospectively to make the register look complete.

10. Service standards

11. Records and evidence

Appendix B readiness. The Ministry may demand copies of all IP assets, or access to them, including registration certificates plus renewal certificates or applications. Where a granted asset exists, this is one of the easier evidence requests in the framework to satisfy, because registering offices issue exactly the documents the Ministry names and technology transfer holds them as a matter of course. The exposure is not the certificate; it is the ownership share, which lives in an agreement rather than a certificate, and which nobody currently has a reason to extract and file alongside the asset. On current assumptions MBRU could produce certificates within 15 working days and would struggle to produce a documented 20% share for every row [IoL to confirm].

12. Risks and controls

13. Performance measures

Watch the disclosure count, not the grant count, in the first two years. Grants lag disclosures by years, and a workflow judged on grants in year one will be judged on decisions taken before it existed. The number of disclosures received is the measure of whether this workflow is working; the number granted is the measure of whether the institution decided to spend, which is a different question owned by different people.

14. Change control