Watch how it works
Log in
Log in app.tacillon.com01
The gap
You already have written procedures. What you do not have is evidence that the person doing the work read them, at the machine, on the day, and confirmed each safety-critical step before moving on.
That gap is where the cost sits. Half of small businesses in Canada spend at least 40 hours a year — about five full business days — informally training a single new employee with no prior work experience.1 The knowledge being transferred lives in one or two experienced heads. When that person retires, quits, or is on another site that morning, the transfer does not happen and the work goes ahead anyway.
It is also what gets examined afterwards. A lawyer writing for Ontario small and medium employers puts it plainly:
Miller Thomson LLP
When an incident occurs, MOL inspectors will not only look at the physical equipment, but also at your written procedures, training records, and how faithfully your workforce follows what is on paper.2
The third item is the one most operations cannot evidence. A binder proves a procedure existed. It does not prove anyone followed it.
02
What a shift looks like with Tacillon
A worker in their second week walks up to a piece of complex machinery. They have a phone. Nothing that follows depends on them remembering anything.
First, four gates must pass
- Which machine? They scan the tag fixed to the unit. Wrong machine, wrong guide — it will not open.
- Are they fit to do it? Declarations about their own condition and their protective equipment. Asking makes it a decision rather than a drift.
- Are they cleared? Authorisation is attested and recorded, so nobody can later say they assumed.
- What state is it in? Baseline condition recorded before anything starts — which protects the worker and the business equally.
Then the work begins, one step at a time
One step fills the screen. No scrolling, no hunting, no reading step nine while performing step seven. Each instruction is audio guided, so eyes and hands stay on the equipment.
Not every step is treated the same, because not every step matters the same:
- Standard The ordinary body of the job. Read it, do it, confirm, move on. No ceremony where none is needed.
- Critical Steps where a mistake is expensive or hard to reverse. Confirmation is deliberate rather than casual.
- Critical-safety Steps where a mistake injures or kills. The worker must affirm a named assertion, captured word for word. There is no next button until it is affirmed.
Reserving the hard stop for the handful of steps that genuinely warrant it is what keeps those stops meaningful. A checklist that treats every step as life-threatening trains people to click through it.
And it has to end somewhere
Abort sits on screen for the whole session. One control stops the job at any point and asks why, from a named list of reasons. Every session ends with a stated outcome — finished, aborted or abandoned. Nothing trails off into ambiguity.
None of this needs a signal. The guide, the gates and the recording all run on the device. A job started in a basement behaves identically to one started in the yard, and the records queue for your archive until signal returns.
03
Why one step at a time
The one-step-at-a-time screen is not a styling choice. It is a response to something measurable about how people now hold attention.
Gloria Mark's group at UC Irvine has tracked how long people stay on a single screen before switching. In 2004 it was about two and a half minutes. It is now an average of 47 seconds, with a median over the last five years of 40 seconds.3
Interruption is not free either. In a controlled study, Gloria Mark and colleagues found that people finish interrupted work faster rather than slower — but pay for it in stress, frustration, time pressure and effort.4
Now put that in an industrial facility. Memorizing a nine-step procedure, in gloves, poor lighting, with interruptions, requires exactly the cognitive skill (sustained, unaided working memory across a sequence) that the preceding evidence says has been degraded over the last two decades — presumably by smartphones, notifications, and fragmented attention patterns documented in the cited sources.
A gated, one-step-at-a-time screen does not ask them to hold the entire sequence. Tacillon holds it. Their place is kept through any interruptions, the step that can hurt them will not let them past, and the record of their work is written as they go rather than reconstructed afterwards.
04
What you are left holding
Everything above produces a record as a by-product of the work being done properly. Each confirmation carries who, what, when, on which machine — and for safety-critical steps, the exact words that were affirmed.
- It leaves the phone. Every entry is copied to your archive as it is written. A lost, stolen or wiped handset does not take your evidence with it. We verified this by deliberately wiping a device and confirming the archive was intact.5
- Nobody can edit it. The archive accepts new records and shows them to you. It is not permitted to change one or remove one — including by your staff, and including by us.
- It can be checked. Each entry is locked in sequence to the one before it, so altering any past entry visibly breaks the chain. The record can be re-verified without the original phone being present, or trusted.
- It is neutral. If a step was skipped, the record says so. A record that only ever showed compliance would be worthless as evidence — which is exactly what makes the compliant sessions mean something.
Said plainly, because it matters: this is risk and liability mitigation, not a court-proof electronic signature. What it gives you is evidence generated at the moment the work happened rather than reconstructed afterwards. That is a materially stronger position than a signed binder. It is not the same thing as a legal instrument, and anyone selling you the stronger claim is overselling it.
There is a second use for the same data. Because stopping is made easy and the reason is captured, you accumulate something most operations have never had: a dated, categorised record of near-misses that never became incidents. Three aborts for the same reason on the same unit in a month is a maintenance signal no informal process would ever have surfaced.
05
What it asks of you
Not the subscription — the effort, which is the thing that actually decides whether a tool like this gets used.
- 01From your expert: a few hours per procedure. Walking the job with a phone, filming short clips, dictating steps. At the machine, in one pass, by the person who already knows it. No writing, no desk work, no document formatting.
- 02From your workers: almost nothing. Open it, scan the tag, follow the screen. There is no training curriculum for using it, because using it is the training.
- 03From you: a decision about scope. Which procedures go first. The right answer is the two or three where a mistake is worst, not the twenty that are merely tedious.
The hard part has never been the software. It is getting your most experienced person to externalise what is in their head. No tool removes that. What changes is that the output is a working procedure rather than a document that goes in a folder, and that the capture takes an afternoon at the machine rather than a writing project nobody finishes.
Log in
Log in app.tacillon.com06
Sources
- Canadian Federation of Independent Business, Canada's Training Ground: How Small Businesses are Building Tomorrow's Workforce, 27 October 2025. cfib-fcei.ca ↩
- Jason Lakhan, Miller Thomson LLP, Top 5 workplace safety risks for Ontario SMEs, 16 June 2026. millerthomson.com ↩
- Shani Murray, Regaining Focus in a World of Digital Distractions, interview with Prof. Gloria Mark, UC Irvine Donald Bren School of Information & Computer Sciences, 26 January 2023. ics.uci.edu ↩
- Gloria Mark, Daniela Gudith and Ulrich Klocke, The Cost of Interrupted Work: More Speed and Stress, Proceedings of CHI 2008, pp. 107–110. dl.acm.org · full text (PDF) ↩
- Sequonyx Inc., internal durability test, July 2026. This is our own test, not an independent audit. ↩