Pilot Discovery Guide
A conversation guide for a Cement Agent Plant Pilot discovery call. It is a static guide for the conversation — not a form and not a promise of service. Nothing is collected or submitted here; it simply helps both sides decide whether a private, advisory-only pilot is worth running.
Advisory only. The pilot is a workflow and knowledge layer — not process control, not advanced process control (APC), and no control-system integration or writeback. It records, structures, and routes; it authorizes nothing. Every decision stays with the appropriate human authority.
1. Conversation purpose
- Decide together whether a private advisory workflow pilot would be useful here.
- Understand how issue intake, shift handover, safety observation, QC, process, and maintenance routing work today.
- Surface where the current approach is inconsistent or hard to hand off — without proposing any control or field change.
2. Role and context
- What is your role, and which plant / line / area and department are we talking about?
- Who owns the workflows in scope, and who would use a pilot?
- Where does current intake / handover / review feel painful or inconsistent?
- What documentation tools do you use today (shift logs, forms, spreadsheets, CMMS notes)?
3. Workflow fit
Which of these advisory workflows would be worth exercising, and why?
- Plant issue intake and triage — is first-touch routing inconsistent today?
- Shift handover from open issues — do open items get lost between shifts?
- Safety observation carry-forward — how are observations recorded and routed now?
- QC out-of-trend review — how is an out-of-trend result reviewed and routed?
- Dust collector trend review.
- Preheater restriction trend review.
- Finish mill ventilation trend review.
4. Source readiness
What could be shared to configure the advisory layer — only what your authority approves?
- Approved SOP / procedure excerpts (or pointers to them).
- OEM / manual references for the chosen workflows.
- Existing shift-log examples (redacted as needed).
- Existing safety observation / near-miss formats.
- Approved plant terminology and area names.
5. Boundaries
- What must remain under human authority (operations, QC, maintenance/reliability, safety, environmental)?
- What must not be captured at all?
- What are the confidentiality and data boundaries?
- Confirm there is no control-system integration and no writeback to plant systems.
6. Pilot success
Qualitative questions to agree on up front — these are evaluation questions, not promised outcomes:
- Is the advisory output useful to the people who would use it?
- Does a structured handoff make routing clearer and more consistent?
- Is missing data made visible early, instead of surfacing late?
- Was the training / onboarding discussion useful for your team?
- Which workflows, if any, would your stakeholders want to take further?
7. Red flags / stop conditions
If any of these come up, the pilot is not the right fit and the request is declined and routed to the appropriate authority:
- A request for the tool to take a control-system action.
- A request for a compliance determination or environmental determination.
- A request for a safety clearance or authorization of any action.
- A request to encode proprietary criteria (limits, setpoints, thresholds, acceptance criteria) without approval from your authority.
- A request to replace qualified personnel or site procedure.
Next steps
To share a concise summary, send the pilot brief. To prepare what is useful to have ready, use the pilot inquiry checklist. On deployment and data handling, see Private Deployment & Data Boundaries, and for ways to bound a first pilot see Pilot Scope Options. To start a conversation, see Pilot Inquiry Contact. For the full overview and the suggested pilot structure, see the pilot page.