One weekend afternoon in early February, friends invited me to their usual meeting place. They wanted to connect sustainable good deeds with community efforts and hoped to have something ready for April’s Baishatun Mazu pilgrimage. Time was short. Could I help?

I took on the technical development. My initial question was straightforward: could the action records collected through forms and LINE groups the previous year become easier to leave behind? Could participants see how far they walked and what they did?

The original article was written before the April 2026 event and recorded development and ideas at that time. Looking back, I need to separate an inference that moved too quickly: a recorded walking distance is not a measurement of actual emissions reductions.

The “400,000 people” in the original title was also an anticipated event size, not a count of platform users. What follows does not turn pre-event expectations into post-event results.

Four-panel comic: What GPS can record; a recorded walking distance is not a measurement of actual emissions reductions. without this walk, how would this person have traveled? An emissions factor needs its supporting details: source, version, units, covered greenhouse gases, and calculation boundary. Both sides of the comparison need explanation. Leave gaps where data is absent and present assumptions separately from observations; Let the footsteps be seen, and let the numbers be questioned.

What I wanted to record was how people participated

What draws me to the Baishatun pilgrimage is how faith becomes a bodily experience. Walking, waiting, departure, the incense ritual, and the return journey bring people into a shared sequence. Kilometers alone cannot tell the whole story.

My friends talked about a platform for good deeds. I wondered whether technology could reduce the burden of filling in forms. The original article described the previous year’s 善足跡 project using forms and groups to collect records. My job was to try connecting those processes through a website and LINE interface.

I collaborated with the 1.5°C Science-Based Reduction Initiative Standards Association (SSBTi), which contributed methodological discussions and a framework while I handled development. This is the collaboration recorded in my original account. It does not establish third-party verification of all the figures or certification by the international SBTi.

Recording and discussing an action has value in itself. That value does not need a large emissions-reduction total to justify it.

What GPS can record, and what it cannot answer

A phone sends coordinates, times, and check-ins so the system can organize segments and distances. But location readings have errors and data can stop arriving. Several check-in points are not a complete route; total attendance is not a denominator for valid data coverage.

Even after verifying a walking record, GPS cannot answer another question: without this walk, how would this person have traveled?

They might have walked anyway, taken a vehicle, or never set out. Those answers make a substantial difference to an avoided-emissions estimate.

I would now separate the displayed figures into three layers:

LayerData neededWhat it can establish
Action recordsSuccessfully received locations, times, check-ins, or self-reportsWhich participation behaviors were recorded.
Organized distanceRules for deduplication, plausibility checks, and missing dataDistance derived under those rules.
Emissions comparison scenarioBaseline transport, distance, occupancy, and factor boundariesA possible emissions difference from a baseline under stated assumptions.

These layers are not interchangeable. A platform sees, at most, data it successfully receives and processes. That does not mean it records every step of 400,000 people.

6,300 tonnes answers an old calculation, not the event’s results

The original calculation assumed 300,000 people walking 100 kilometers each, multiplied by 0.21 kilograms of CO₂ per kilometer, to obtain 6,300 tonnes.

The multiplication itself is correct:

300,000 × 100 × 0.21 = 6,300,000 kilograms = 6,300 tonnes.

This only reproduces the old calculation. It is neither a factor endorsed in this revision nor a new estimate of event outcomes. The original 0.21 parameter differs from later project documentation and lacked a traceable official factor entry and version. I withdraw the description of it as an “announced Ministry of Environment value.”

The larger problem was treating kilometers walked as the same number of car-kilometers avoided. That silently assumes every person would otherwise drive alone over the same distance, and that the vehicle trip was actually canceled.

If someone would have walked anyway, those kilometers cannot be counted as transport substitution caused by the system. If several people would have shared one car, its whole emissions cannot be assigned to each passenger in turn. If the driving and walking routes differ, the distances must also be checked.

I have therefore removed the forest-park equivalence and the claim of “6,300 tonnes avoided.” Substituting a more precise factor cannot supply transport behavior that was never investigated.

To estimate emissions, describe both sides of the comparison

Even when a participant reports they would otherwise have taken a vehicle, further questions remain: what kind, how many occupants, which trip was replaced, and would the vehicle have run anyway for other passengers? These determine whether vehicle-kilometers or passenger-kilometers apply and how emissions are allocated.

An emissions factor needs its supporting details: source, version, units, covered greenhouse gases, and calculation boundary. Kilograms of CO₂ and kilograms of CO₂e cannot be used interchangeably without explanation. The latter expresses the included greenhouse gases as carbon-dioxide equivalents; its coverage is not automatically the same as the former’s.

A pilgrimage also includes more than that walking segment. Whether travel to meeting points, shuttles, accommodation, and supplies belong in the calculation depends on the question. An estimate limited to one transport-substitution scenario should be named accordingly, not called the net reduction or full carbon footprint of the whole event.

Both sides of the comparison need explanation. Leave gaps where data is absent and present assumptions separately from observations, so readers can change the conditions and see how the result changes.

Four weeks of development: the human interface still interests me most

Over roughly four weeks, I connected phone GPS check-ins, backend APIs, LINE integration, a database, questionnaires, and a dashboard. Astro handled the frontend, with Cloudflare Workers, D1, and KV supporting backend needs. Participants would not need another app.

Edge computing was an engineering choice. It did not make phone location readings precise or prevent mobile-network interruptions along the route. The original load-test figures also lacked conditions and reports readers could reproduce, so I no longer use them to promise capacity at the event.

Claude helped accelerate architecture drafts, code, and test scripts. I still had to decide why things were built that way, what needed checking, and pay the tool and token costs. Speed gave the idea a chance to take shape; acceptance checks told me what had actually been achieved.

My favorite detail was the nine pilgrim ranks inspired by A Record of a Mortal’s Journey to Immortality: from Qi Refining, Foundation Establishment, Core Formation, and Nascent Soul through to Ascension, each with its own figurine.

Figurines for nine pilgrim ranks, from Qi Refining to Ascension

Turning distance and participation records into progress gives an abstract accumulation a visible form. But a higher rank does not mean an estimate has passed carbon-accounting verification. Labels should say whether a figure is recorded distance or an emissions comparison under a transport assumption.

A sustainability story should let readers check its numbers

GHG Protocol guidance requires avoided-emissions estimates to explain conditions including baselines, boundaries, methods, and data sources. They must be reported separately from corporate Scope 1, 2, and 3 inventories and cannot simply be deducted from Scope 3 emissions.

A participation record can help describe a sustainability activity. That does not make it a figure a company can directly offset, nor does multiplying by a factor automatically create carbon credits. Formal claims need the conditions of the relevant methodology and system.

I hope the next stage makes three things clearer: how many people and which periods the records actually cover, what assumptions the emissions comparison uses, and where data remains insufficient. Public sharing should also use only necessary aggregates, rather than make a person’s full movements the price of participation. These are directions for improvement, not features verified in this article.

Returning to that afternoon’s invitation, I still think the attempt was worthwhile. Turning a wish into an interface where people could check in and leave a record connected traditional ritual with contemporary questions.

Let the footsteps be seen, and let the numbers be questioned. Records made that way have a better chance of accompanying the project further.