# BattBox Pilot Delivery Plan

**Objective:** Decide, with evidence, whether BattBox can be operated safely, lawfully, usefully and economically at marketplace parcel locations.  
**Planning horizon:** 46 weeks to complete six-site pilot, assuming each gate passes  
**Pilot live period:** 16 weeks  
**Status:** Indicative R&D plan

## 1. Delivery principles

- Safety, legal route and insurance are gates, not parallel paperwork to finish after installation.
- Start with the smallest accepted battery envelope and expand only through evidence.
- Test the service and downstream route as deeply as the kiosk.
- Use supervised, indoor beta sites before any unsupervised or outdoor operation.
- Record assumptions, owners and decisions in a controlled log.
- Do not publicise environmental outcomes before the evidence and claims method are approved.

## 2. Governance

### Steering group

- Marketplace executive sponsor — accountable for programme value and decisions.
- Product director — accountable for consumer and platform product.
- Safety / engineering lead — design authority for the physical system.
- Battery Compliance Scheme / treatment lead — accountable for accepted route and evidence.
- Operations lead — accountable for field service, QA and incidents.
- Legal / compliance lead — regulatory role map and partner terms.
- Data Protection Officer delegate — DPIA and data governance.
- Independent assurance adviser — challenges evidence at gates.

### Cadence

- Weekly delivery review.
- Fortnightly hazard and decision-control board.
- Monthly sponsor steering group.
- Immediate critical-event escalation.
- Formal evidence review before every gate.

## 3. Work breakdown structure

### WP1 — user and market discovery

- Interview consumers who store, return and incorrectly discard batteries.
- Observe existing retailer take-back points.
- Test comprehension of accepted/refused examples.
- Map non-smartphone and assisted journeys.
- Test incentive levels and unintended behaviours.
- Validate host-site operational concerns.

### WP2 — regulatory, insurance and IP

- Map producer, distributor, BCS, collection-point, waste-holder, carrier and treatment roles.
- Review each UK nation's waste and environmental requirements.
- Confirm product conformity, electrical, radio, metrology and accessibility work.
- Engage insurer and host property/fire teams.
- Commission patentability and freedom-to-operate search.
- Search trademark and design-right position for final brand and enclosure.

### WP3 — safety and mechanical engineering

- Define battery acceptance envelope and foreseeable misuse.
- Select pouch/tab concept through material and treatment compatibility tests.
- Design scale, interlock, short deposit path and cassette.
- Define sensing, safe states, service lockout and emergency access.
- Build instrumented bench rig.
- Execute destructive and representative fire tests under competent control.

### WP4 — digital product and data

- Marketplace sandbox: token, locator, consent, wallet and notifications.
- Edge state machine and secure device management.
- Deposit, pouch, cassette, QA, collection and treatment ledger.
- Rules engine and case-management console.
- Privacy, security and fraud testing.
- Controlled evidence exports and partner dashboards.

### WP5 — operations and treatment

- Contract BCS, carrier, QA site and approved treatment route.
- Define cassette exchange and maximum dwell/fill rules.
- Write acceptance, jam, quarantine, emergency and incident SOPs.
- Train and assess technicians and reviewers.
- Reconcile scales at site, QA receipt and treatment.
- Run tabletop and live drills.

### WP6 — pilot research and evaluation

- Establish baseline behaviour and comparison sites where feasible.
- Instrument funnel, uptime, quality and operating cost.
- Conduct observed-use and accessibility research.
- Independently review safety, privacy and claims evidence.
- Produce scale, redesign or stop recommendation.

## 4. Schedule and gates

### Gate 0 — feasibility and operating model, weeks 0–6

**Deliverables**

- Product brief and success thresholds.
- Sponsor and partner RACI.
- Signed Phase 1 acceptance envelope in principle.
- Legal/regulatory issue map.
- Preliminary fire strategy and test brief.
- Site shortlist and interface survey.
- Preliminary threat model and DPIA screen.
- Reward and unit-economics model.
- IP/FTO professional search brief.

**Decision:** proceed to bench alpha only if a viable legal/treatment route, insurer engagement, sponsor funding and no fatal safety issue exist.

### Gate 1 — bench alpha, weeks 7–18

**Deliverables**

- Instrumented deposit-path rig.
- Pouch/tab prototypes and test results.
- Edge controller and simulated marketplace flow.
- Device/platform event ledger.
- Operator console and QA simulation.
- Mechanical, electrical, metrology, misuse and preliminary thermal evidence.
- Draft installation, service and emergency procedures.

**Decision:** proceed to supervised beta only after independent technical review and closure of intolerable hazards.

### Gate 2 — supervised beta, weeks 19–30

**Configuration**

- Two indoor, staffed, low-volume locations.
- Published restricted hours.
- Trained operator available; no untrained host handling.
- Low pouch and cassette limits.
- Capped test rewards.
- Daily inspection and evidence review.

**Deliverables**

- Observed completion and comprehension results.
- Clean-deposit quality estimate.
- Service time and incident findings.
- Accessibility remediation.
- Updated hazard log, SOPs and commercial model.

**Decision:** proceed to unsupervised pilot only with zero critical events, at least 90% clean deposits and approved operational evidence.

### Gate 3 — six-site pilot, weeks 31–46

**Configuration**

- Six sites in two UK regions.
- Urban and suburban demand mix.
- Indoor first; outdoor only where verified.
- Up to 1,500 opted-in customers.
- 16 live weeks.
- Remote health/fill monitoring and dynamic collections.
- Real, capped loyalty rewards.
- Independent evaluation.

**Decision:** scale, redesign or stop against the pre-agreed scorecard.

## 5. RACI summary

Key: **A** accountable, **R** responsible, **C** consulted, **I** informed.

| Deliverable | Marketplace | BattBox product | Engineer/manufacturer | BCS/treatment | Operator | Legal/DPO | Independent assurance |
|---|---|---|---|---|---|---|---|
| Product requirements | A | R | C | C | C | C | I |
| Acceptance envelope | C | C | R | A/R | R | C | C |
| Fire strategy | I | C | A/R | C | C | I | C |
| App and wallet | A/R | R | I | I | I | C | I |
| Hardware design | I | C | A/R | C | C | I | C |
| Waste/treatment route | I | C | C | A/R | R | C | I |
| Field operations | I | C | C | C | A/R | I | C |
| DPIA and data terms | C | R | C | C | C | A/R | C |
| Gate evidence pack | A | R | R | R | R | R | C |
| Scale decision | A | R | C | C | C | C | C |

## 6. Resource plan

Core team estimate during Gate 1–3:

- Product/programme lead: 1.0 FTE.
- Hardware systems/safety engineer: 1.0 FTE.
- Mechanical/electrical engineering supplier: project team.
- Firmware engineer: 0.5–1.0 FTE.
- Backend/full-stack engineers: 1.5–2.0 FTE.
- Product designer/researcher: 0.5–0.8 FTE.
- Operations lead: 0.5 FTE rising to 1.0 during pilot.
- Safety, fire, waste, legal, privacy, accessibility, cyber and IP specialists: scoped engagements.
- Field technicians/reviewers: capacity driven by site and deposit volumes.

## 7. Budget model

The pilot budget covers discovery, specialist assurance, engineering, platform integration, supervised beta, site hardware, live operations, treatment, rewards, evaluation and contingency. Values are protected because they are active sponsor-negotiation assumptions rather than public supplier quotations.

Open the protected commercial room for the current pricing and unit-economics model:

https://battbox.co.uk/commercial

The master budget must avoid double-counting reusable Gate 0 and Gate 1 assets. All values exclude VAT until quotations and the partner role map are complete.

## 8. Pilot instrumentation

### Safety

- Sensor alerts by type and duration.
- Shutter, jam, retrieval and service-lock events.
- Refused-item and emergency-help selections.
- Incidents, near misses and corrective actions.

### Funnel

- Locator view -> token issue -> scan -> screen -> kit -> valid weight -> deposit -> verified reward.
- Duration, cancellation and failure reason at each state.
- First-time versus repeat experience.
- Accessibility route usage.

### Quality and fraud

- Accepted pouch count and net weight.
- Sample rate and contamination categories.
- Required terminal isolation accuracy.
- Weight discrepancy by site and instrument.
- Account and device anomaly patterns.
- Appeals and overturned decisions.

### Operations

- Fill, mass, dwell and temperature history.
- Travel, service and cassette-swap time.
- Availability and fault mean-time-to-recover.
- QA labour and quarantine events.
- Treatment receipt and mass reconciliation.

### Commercial

- Fully loaded cost per accepted kg and verified deposit.
- Reward cost per new and retained participant.
- Partner funding split.
- Incremental app sessions and campaign redemption without overstating causality.

## 9. Scale-decision scorecard

The steering group agrees numerical thresholds at Gate 0. Minimum suggested thresholds:

- Zero critical safety events.
- At least 95% published-hours availability.
- At least 90% clean deposits by sampled pouch count.
- At least 70% valid-token flow completion.
- Median experienced-user flow under 75 seconds.
- 100% cassette transfer and treatment records reconciled.
- Collections before 75% fill or maximum dwell time.
- At least 80% correct preparation in observed research.
- No unresolved high-severity privacy or security findings.
- Fully loaded cost per accepted kg below the sponsor's agreed ceiling.

Passing average metrics does not override a critical safety, legal, insurance or ethical failure.

## 10. RAID log starter

| Type | Item | Owner | Trigger / decision |
|---|---|---|---|
| Risk | Thermal event or propagation | Safety lead | Any alarm/test failure pauses affected configuration |
| Risk | Damaged battery deposited | Operations | Quarantine SOP and acceptance redesign threshold |
| Risk | Rewards increase unsafe collection | Product | Cap, messaging and incentive change |
| Risk | Transport route rejects cassette/pouches | BCS lead | No beta until packaging route signed |
| Assumption | Marketplace can issue one-time token | Marketplace | Sandbox proof by Gate 1 sprint 2 |
| Assumption | Six sites provide suitable power/separation | Host lead | Site survey before detailed manufacture |
| Issue | Exact metrology duty unknown | Legal | Advice before reward beta |
| Dependency | Approved treatment evidence interface | BCS lead | Schema agreed before platform freeze |
| Decision | Indoor versus outdoor pilot | Safety board | Evidence review at Gate 2 |

## 11. Communications and claims

- Use “concept,” “test” and “pilot” until safety and compliance evidence supports stronger wording.
- Attribute the 2026 fire figures to the ESA report as covered by the source.
- Do not call a collected pouch recycled.
- Publish accepted, rejected, collected, received and treated figures with definitions.
- Record methodology changes and do not back-cast impact without explanation.
- Prepare incident communications and a regulator/host notification matrix before launch.

## 12. Exit criteria

### Scale

All mandatory gates pass, economics meet the sponsor threshold and partners sign a controlled scale plan.

### Redesign

No critical unresolved risk, but one or more experience, quality, availability or commercial thresholds need a bounded redesign and retest.

### Stop

Any uncontrolled intolerable safety risk, unavailable lawful/insured route, unacceptable privacy impact, treatment partner withdrawal or structurally uneconomic service model.

---

This is a planning document, not a public-deployment approval or supplier commitment.
