To seek approval for Annual Camp 2027 training to be delivered by standing Subject Matter Expert (SME) teams organised by activity stream, with cadre staff retained to command, move and care for their cadets.

That the CTO approves the SME stream delivery model for AC27 (Option 2, para 7), and authorises a design team to be stood up by 30 November 2026 to produce the stream plans against the timeline at para 10.

We already run a rotation matrix. We do not yet put a permanent team on each activity.

Reference A already cycles 9 cadres through 5 principal activity streams across 12 training half-days. Read by row it is Company training. Read by column it is an SME stream plan — already drawn, already deconflicted, already approved.

This proposal does not change the timetable, the activities, or the cadre allocations. It changes who owns the columns. The structural work is largely done; what is missing is a permanent team behind each activity. This is shown at Figure 1 of the covering summary.

Until approximately two years ago the County trained in Companies — A, B, C and S — each holding a mix of Junior 1-Star, Senior 1-Star and 2-Star cadres, with 3 and 4-Star delivered by S Coy. The County has since moved to a cadre structure: one cadre per Company (AC26: A = Junior 1-Star, B = 2-Star, C = Senior 1-Star, S = 3/4-Star).

The intent of that change was sound. My assessment is that the concept was right and the implementation was incomplete: we adopted cadre allocation without adopting cadre delivery. Companies still approach camp as Companies — each planning, resourcing and delivering the full syllabus for its own cadre, in parallel with every other Company.

This proposal is the missing half of a change already made, not a new direction.

AC26 delivered its training. It did so at a cost that is not sustainable and not, in my judgement, a function of how hard anyone worked.

5a. Staff were working until approximately 2200 to prepare the following day. This is a morale, retention and health risk. Its cause is structural, not individual: nine cadres each re-planned and re-drew stores every evening, off the back of a nightly meeting, against a QM department receiving requirements through several uncoordinated channels. The work was duplicated nine times and could only be done at night, because it could only be done once the day's training had finished.

5b. G4 requirements did not consistently reach the point of issue. Requirements were submitted in advance at cadre level (Ref C) and again nightly during camp; some did not arrive. With nine cadres each raising daily demands, the number of transactions is high and each one is a point of failure.

5c. Duplication is already occurring across counties. On 11 of the 12 training half-days, at least one activity stream is running cadres from both Gloucestershire and B&CI simultaneously (Ref B, para 6) — the same activity, on the same day, with separate staffs, separate lesson plans and separate stores demands.

5d. Consistency of training cannot currently be demonstrated. Where the same syllabus is delivered by whichever staff are available to each cadre, we have no mechanism that assures a common standard or a common assessment threshold across cadres. I do not assert that standards varied at AC26 — I assert that we could not prove they did not.

Annex A re-reads Reference A by column. Every figure is reproducible from the block programme; nothing has been altered.

The three points I would draw out:

• Para 5a is solved at root. Stream ownership moves planning from during camp to before camp, and from nine parallel efforts to one owner per subject. A nightly stores transaction becomes a single draw and a single return.
• Para 5b is reduced by design. "All bergens and CTRs" becomes one conversation with the FTX stream lead, not nine separate demands.
• Para 5d is answered directly. One team owns each subject, therefore one team owns the standard and the sign-off. Consistency stops being a hope and becomes a structure.

The secondary prize is adult development. A junior adult currently learns from whoever they happen to be stood next to. Under this model they spend a week embedded with the County's strongest instructors in one subject — seeing training planned properly and delivered well, and contributing to delivery where the stream lead judges them ready. That is a succession and retention mechanism we do not currently have.

The exception must be stated plainly. Weapons and ammunition cannot be held for the week. They are drawn and returned daily, against the established armoury accounting and security requirements, and this proposal does not seek to change that. The Ranges stream therefore runs a daily armoury cycle, and applies the draw-once principle only to its range stores — targetry, ancillaries, butt party equipment and consumables.

That exception is smaller than it first appears, because weapons scale to the firing point, not to cadet strength. Worked example — 60 cadets on a 25m barrack range with 6 lanes gives 10 details of 6. Six weapons are on the firing point. Each concurrent practice bay pairs two details on six weapons: one cadet handling, one watching for errors and acting as assistant cocker.

The requirement is therefore 18–30 weapons, not 60. This yields three benefits:

1. A smaller daily armoury cycle — the one transaction the model cannot avoid is materially reduced.
2. Weapons released to concurrent training — the balance stays available to Fieldcraft, SAA or Drill stands running at the same time, which is currently a routine point of contention.
3. Better shooting and fewer range errors — every cadet revises the marksmanship principles unhurried (rifle lessons 5–9, practice period 3) immediately before firing, with a second cadet observing. Errors are corrected in the practice bay rather than on the firing point.

The practice periods are run by cadre staff under the stream lead's direction — which is also the clearest answer to what cadre staff do while the stream owns the stand (Risk R2).

Option 1 — Retain cadre delivery (status quo). For: no change required; known quantity. Against: all of para 5 recurs. The 2200 finishes are structural and will return. Rejected.

Option 2 — SME stream delivery (RECOMMENDED). Standing teams own FTX, Ranges, AT, Nav/Exped and the decomposed ACS stands. Each team plans its subject once, months out, and holds its stores for the week (weapons and ammunition excepted — para 6a). Cadre staff — one senior adult and a cohort of junior adults — remain with their cadre throughout, commanding the cadets and moving them between stands. For: addresses paras 5a–5d; teams demonstrably employed (para 6); stores scale and transaction count fall sharply; training quality and consistency owned by subject; junior adult development built in. Against: requires the design work at para 9 to be completed before camp; concentrates qualifications (Risk R1); the Ranges stream retains a daily armoury cycle (para 6a). Recommended.

Option 3 — Partial adoption (fallback). Streams stood up for FTX and Ranges/SAA only — the highest G4 load and highest qualification burden — with the remainder cadre-delivered. For: smaller step; reversible. Against: retains parallel planning for the majority of the programme, so para 5a is only partly relieved. ACS, the largest block and the weakest area (para 9), is left untouched. Offered as a fallback if Option 2 is judged too large a step for a single year. It is a slower route to the same place, not a different answer.

Figure 4 states the division of responsibility. Figures 5 and 6 draw it, because the question the model must answer is not who owns the subject — that is the easy part — but who commands whom, at every level, on every day of the week.

The cadre structure does not change (Figure 5). A Company HQ administers its cadre. A Lead Instructor owns that cadre's training. Platoons are the manageable block. The section is where an adult stands in front of cadets, with a cadet NCO beneath them. All nine cadres are commanded through those same four levels, from the 70-strong 1-Star cadres down to the 4-Star Advanced cadre of ten — which is the same tree, one section deep. Nothing in this proposal alters any part of that.

The stream team is built on the same four levels (Figure 6). A stream lead sits where a Company HQ sits; a stand or a phase sits where a platoon sits; an instructor stands where a squad instructor stands. We are not introducing a new kind of organisation, and we are not asking anyone to command in an unfamiliar shape. We are pointing a familiar one at a subject instead of at a cadre.

That is the practical case for the model as much as the theoretical one:

• Span of command stays inside what one person can hold. No appointment in either figure commands more than four or five subordinate elements. This is the reason the Coy / Pl / Sect structure has survived, and it is why the stream teams are drawn to it.
• Every level has one named owner. Where a Lead Instructor owns a cadre's training, a stream lead owns a subject's. Consistency (para 5d) and sign-off (R4) attach to a named person, at a named level, in both trees.
• Named deputies are structural, not administrative. R1 asks for a qualified deputy against every safety-critical appointment. Figure 6 shows where each one sits, which is how that risk is actually retired.
• The two trees join, they do not merge. At the foot of each stream tree is a dashed band: cadre staff attach to the stream for the duration of a stand and remain in their own chain of command throughout. Attached, never commanded. That is the whole of the safeguarding answer at R2, drawn rather than asserted.

Two caveats on the figures. Adult strengths in Figure 5 are illustrative ratios, not a manning statement — sections of about ten, three to a platoon. The appointments inside each stream team in Figure 6 are my first draft for the design team to correct; the four levels, not the job titles, are what is being proposed.

ACS peaks at 416 cadets across 8 cadres simultaneously and carries 315 across 5 cadres on the Sunday. It cannot be a single SME team, because ACS is not a subject — it is the residual bucket into which Drill, CIS, Skill at Arms, First Aid/EFAW and expedition theory are swept.

It must be decomposed into named subject stands, each with an owner, a lesson plan and a stores line. This is the largest single piece of work the model requires.

It is also the clearest diagnosis of AC26: ACS is precisely where cadre staff had no pre-planned product to deliver, and therefore where the improvisation and the late nights concentrated. On the Sunday, five separate staffs delivered five versions of the same syllabus to 315 cadets at the same time. Decomposed into stands, that becomes one team per subject delivering to a rotation.

AC26's block programme was issued on 3 Mar 26. SME teams need it earlier, because their value comes from planning far out — that is the whole mechanism by which the work moves out of camp evenings.

Two dates carry the argument: a consolidated G4 bid in April rather than June/July, and stores packed by stream rather than by cadre.

1. Approval of Option 2 — SME stream delivery for AC27.
2. A design team stood up by 30 Nov 26, with stream leads named.
3. A decision on R5 — whether streams are constituted at Battalion level with B&CI, or at County level for Glos-led cadres only.

I am willing to lead or sit on the design team, and to own the FTX stream plan as a worked first example. The S Coy AC26 FTX and week stores plans were produced on exactly this basis and are available as evidence of the product a stream lead would deliver.

The training at AC26 was delivered, and the people delivering it worked extremely hard to achieve that. My concern is that we achieved it by absorbing the cost personally, and that the cost falls on the same people every year.

The change I am proposing is smaller than it first appears. The rotation already exists in your block programme. What it lacks is a permanent team behind each column — and once those teams exist, the planning moves out of camp evenings and into the spring, where it belongs.

SSgt B McCullough Salamanca Coy, Gloucestershire ACF

Annexes: - A. Activity Stream Matrix — worked example using AC26 figures - B. AC26 Supporting Evidence