R11 — Funding and partnerships (useful without affiliation)

Status: Wave 1 research recommendation
Audience: Programme leads; anyone drafting MoUs, grant bids, or public claim language
Non-goals: Invented impact metrics; assuming TfL approval or endorsement; committing to specific funders by name as decided partners

---

1. Decision question

How should TUNES relate to operators, authorities, universities, charities, and funders so that:

1. Passenger and research value grow without implied endorsement
2. Evidence offered to professionals is useful and honest
3. Funding does not capture open data or silence inconvenient findings
4. Ethics and legal timing match campaign ambition

Accepted constraints:

• TUNES is independent; not operated, approved, or endorsed by Transport for London (project-charter)
• Professional practice notes (e.g. TfL R3291 via FOI) may inform methodology as references, not affiliation or redistributable data
• Open releases owned by tunes; apps in sibling repos (repos.md)
• Prefer derived over raw; refuse unexplained health/risk scores until justified (01-assumptions)

---

2. Relationship model: adjacent, not affiliated

Think in lanes, not a single “partner” label.

Decision: Default public language is independent citizen / research programme. Any deeper relationship needs a written scope that preserves non-affiliation and open-release rights (R10).

---

3. What evidence is useful to operators (without being them)

Operators already run (or can procure) Class-rated surveys. TUNES should not sell itself as a cheaper Class 1 substitute. Useful adjacent evidence tends to be:

1. Repeated in-service coverage across times of day, loads, and carriage positions that one-off engineering runs rarely sample densely
2. Passenger-reported problem recurrence — where subjective complaints cluster alongside (not blended into) objective section metrics
3. Spatial/temporal hotspots as hypotheses — ranked sections for follow-up professional measurement, not regulatory proof
4. Method transparency — schema, quality tiers, uncertainty, exclusions, pipeline versions so staff can dismiss or trust knowingly
5. Portable comparison units — station-to-station sections with duration, aligned with practice they already recognise in methodology notes

Useful packaging for a professional reader:

• Versioned open release (R10) with clear limitations
• Quality-filtered subset (e.g. higher tiers only) as an optional export
• Separation of objective metrics vs perception
• Explicit “not engineering-grade / not regulatory” framing

Do not invent coverage percentages or dB deltas in outreach. Cite only measured releases once they exist.

---

4. What makes operators and authorities cautious

Anticipate friction; design partnership asks accordingly.

Decision: Do not seek endorsement as a launch gate. Seek clarity of independence first; optional technical dialogue second; formal MoU only when mutual obligations are concrete.

---

5. Academic partners — value and sequencing

Universities add:

• Ethics board pathways and research-governance literacy
• Calibration / device-limit experiments (R3, R4) with proper protocols
• Statistical and acoustic review of claim language
• Citation and archival norms for open datasets
• Possible access to Class-rated reference gear for comparison studies (still not making TUNES Class 1)

Sequencing recommendation:

1. Internal methodology + privacy defaults documented
2. Ethics advice on whether public beta campaigns need review (see §7)
3. Pilot-scale collaboration on calibration / validation experiments
4. Publication of methods + versioned data under TUNES licences
5. Only then larger consortium grants that name TUNES as open infrastructure

Avoid exclusive IP or embargos that prevent tunes from releasing L2 aggregate packages.

---

6. Charities and community groups

High alignment with charter public value (parents, sensory-sensitive travellers). Useful roles:

• Co-design of subjective instruments (R9) and accessibility of the recorder/map
• Review of claim language for harm (over-reassurance or alarm)
• Recruitment ethics for pilots (consent clarity, no coercion)

Keep measurement science ownership in tunes; community partners advise, they do not silently redefine tiers.

---

7. Ethics approval — timing

Open question in 01: whether ethics review is required before public beta campaigns.

Working guidance (not a substitute for institutional advice):

Decision: Treat ethics + DPIA timing as a launch gate for public recruitment, not for writing methodology docs. Prefer university or qualified privacy counsel over informal “we’ll be fine.”

If no academic partner yet, still commission privacy/legal review before wide collection; controller identity follows ADR-013.

---

8. Funding compatible with open data independence

Compatible patterns

• Grants that require open methods and open data
• Funding for infrastructure (pipeline, schema, accessibility) without editorial control of findings
• University seed that pays student time under TUNES licences
• Civic tech funds that accept independent branding

Incompatible or high-risk patterns

• Sponsor approval rights over publications or release contents
• Exclusive commercial licence that blocks CC BY / CDLA-style reuse
• Co-branding that implies operator endorsement as a condition of payment
• Embargos that prevent publishing negative or null results
• Bundling non-open operator reports into “the” dataset

Safeguards to write into bids and MoUs

1. TUNES retains right to publish versioned releases under the programme licence (R10 / licences.md)
2. Non-affiliation clause: funder/operator logos only with written permission; no endorsement language
3. Scientific disputes resolve via ADRs in tunes, not sponsor memo
4. Privacy defaults (derived-first) are non-negotiable without a public ADR
5. Claim-language limits remain in force

Provisional licence direction (MIT code + CC BY docs/data) supports independence; final text still needs legal review under ADR-012. Prefer avoiding non-commercial-only data licences if they later conflict with mixed academic–civic reuse — confirm with counsel.

---

9. GLA and city-level actors

GLA / city innovation or environment teams may care about exposure evidence and inclusive transport. Same rules as TfL lane:

• Useful as audience and possible funder, not as owner of the science
• City open-data portals may mirror tunes releases; canonical citation stays with tunes
• Policy interest ≠ methodological endorsement

Do not invent citywide prevalence claims from a dense London-first pilot.

---

10. Partnership sequencing (recommended)

Phase 0  Docs, charter, non-affiliation, privacy defaults
Phase 1  Legal + ethics/DPIA advice; no mass recruitment yet
Phase 2  Small pilot; academic calibration experiments; first internal derived sets
Phase 3  Versioned public release (R10); self-serve operator/academic reading
Phase 4  Optional MoUs: data use talks, accessibility collabs, grants
Phase 5  Only if mutual value clear: deeper technical exchange (still no endorsement claim)

Seeking endorsement or formal operator partnership before Phase 3 invites either capture or rejection based on an immature method. Open reproducible releases are the primary “handshake.”

---

11. Communication checklist (every external touch)

• State independence / non-affiliation
• State survey-grade / not Class 1 substitute
• Point to schema, quality tiers, uncertainty (or “not yet released”)
• No invented statistics
• No blending objective + subjective into one score
• No redistribution of non-open FOI reports as open data
• Funding asks preserve open-release rights

---

12. Decisions deferred to Wave 2

• Named target funders and bid calendar
• Whether a light-touch advisory board (academic + civic) is worth the overhead
• Exact MoU template clauses (legal)
• Whether any operator data-sharing agreement is ever needed (default: no — we publish ours)

---

Recommendation

Treat operators and the GLA as informed audiences and optional later collaborators, never as implied endorsers. Lead with versioned open evidence that is useful as in-service, multi-condition hypothesis generation, not as regulatory proof. Prioritise academic partners for ethics pathways and calibration experiments; charities for accessibility and claim-language review. Gate public recruitment on ethics/DPIA advice. Accept only funding that preserves open releases, non-affiliation, and editorial independence of methodology and results.

Confidence: High for non-affiliation sequencing and “useful without substituting Class 1.” Medium for exact ethics trigger points (institution-dependent). Low for specific funder fit until bids are drafted with legal review.

Depends on experiment/legal/user-test? Legal (licences, controller, MoU clauses). Ethics/DPIA before public campaigns. User tests (R8/R9) strengthen charity/accessibility partnerships. Calibration experiments (R3/R4) strengthen academic and operator credibility — they do not require endorsement.

Links to related docs:

• ../governance/project-charter.md
• ../governance/scope-statement.md
• ../governance/licences.md
• ../governance/repos.md
• ../01-assumptions-and-open-questions.md
• ../02-research-map.md
• ../acoustic-survey-methodology.md
• 10-open-data-reproducibility.md
• 05-privacy-ethics.md
• 01-citizen-science.md