Phased pilots and transition review

1. Identify the change being tested

A transition might add a greenhouse to a field-based supply, introduce a fermented ingredient or share processing and storage between producers. Begin with the problem and useful food function, rather than treating adoption of a new process as the goal. Identify the people affected, current service, intended improvement and constraints.

Record which part of the system will change and what must keep working during the trial. Existing supply, trained staff and maintenance arrangements have value. Removing them before an alternative is demonstrated can create an avoidable gap. Food-system transitions need both production evidence and evidence about access, resources and supporting services.

2. Different stages answer different questions

A laboratory result may demonstrate that one organism or process can produce an ingredient under specified conditions. It does not automatically establish a safe finished food, reliable weekly output, affordable operation or a commercial system. A small operating pilot can answer some of those additional questions, while others need longer observation or a different test.

For each stage, write what is demonstrated, what is assumed and what remains unknown. Track feedstocks, nutrients, energy, water, waste, processing and infrastructure at the actual tested boundary. An emerging process should remain labelled emerging when broader performance is still projected. Detailed production mechanisms and food-safety validation belong in their primary topics.

3. Make the pilot reviewable

A pilot plan names its useful output, period, participants, responsibilities and comparison basis. It identifies how measurements will be collected and which observations would support continuation, revision or stopping. A production milestone is separate from the safety and lawful-operation conditions that require competent, process-specific assessment.

Limit commitments to what the evidence justifies. A reversible trial might use temporary arrangements while preserving existing supply, but reversibility has costs and limits. Ask what happens if the pilot stops, who bears extra work and how participant data and feedback are handled. Neither a small sample nor a temporary installation removes those responsibilities.

4. Use actual output in a resource ratio

In an invented one-week pilot, the target is 40 kg of a specified usable product delivered at a defined receiving point. Actual delivered output is 30 kg, with the same specification. A stated meter records 24 kWh for the included pilot equipment over that week. The output is 30/40 = 75% of the quantity target, with a 10 kg gap.

The measured electricity intensity within this boundary is 24 kWh / 30 kg = 0.80 kWh/kg. Dividing by the planned 40 kg gives 0.60 kWh/kg, but that is not the intensity of the actual pilot output. The account excludes upstream inputs, building services not on the meter, distribution beyond the receiving point and other environmental impacts.

The same product specification is an explicit case assumption, not safety clearance. One week provides no annual reliability estimate. The result can guide questions about causes, operating conditions and further tests; it does not prove a commercial cost or a universal improvement.

5. Review effects, not only activity

Compare actual results with the recorded goals and baseline. A before-and-after change can be influenced by weather, demand, prices or selection of participants. It is descriptive evidence unless an appropriate comparison and its assumptions support a causal claim. Feedback can identify practical barriers that an output total misses.

A missed target is a reason to investigate, not automatic proof that the whole method is unsuitable. A met target is also limited evidence: examine rejected output, resource use, costs, interruption recovery and effects on other actors. Record revisions openly rather than changing the original goal after seeing the result.

6. Decide what the next stage needs

Wider adoption may require different equipment, staffing, utility capacity, maintenance, logistics and finance. Multiplying a pilot's output does not demonstrate that these dependencies will scale proportionally. Check the conditions under which the pilot worked and the new conditions it would encounter.

Choose a bounded next step: repeat under different conditions, resolve a specific bottleneck, compare an alternative or stop the proposed change. Keep the evidence and remaining uncertainty visible. A transition is progress when the intended food function and supporting system improve under tested conditions, not simply when a larger facility is installed.

Worked examples

Check your understanding

1. Does a successful laboratory ingredient demonstration prove affordable continuous commercial food production?

   Show answer and reasoning

   No.

   A component result leaves additional production, safety, resource, infrastructure and access questions.

2. Can a pilot's measured electricity be divided by planned output to report actual intensity?

   Show answer and reasoning

   No.

   Use the actual useful output at the matching boundary and period, or clearly label a different hypothetical calculation.

3. What evidence should accompany a decision to expand a pilot?

   Show answer and reasoning

   The tested conditions, actual outcomes, resource and service dependencies, comparison limits and unresolved risks.

   A larger commitment needs evidence about the additional conditions it introduces, not just a multiplication of pilot output.

Key ideas

• Define the food function and preserve necessary services during a trial.
• Separate laboratory demonstration, operating pilot and projected wider performance.
• Use actual output and an explicit boundary in measured resource ratios.
• Keep original milestones, comparison limits and the reasons for revisions visible.