5 ways to automate near-miss reports
I'm Liam Jones, founder of Pilla and a qualified management consultant. I've helped hundreds of businesses set up workflows, and in this article I'm going to show you five real examples of how to set up your near-miss reports.
I'll start with the simplest version, then show you one addition at a time, so you can pick the options your business needs. You can open up each template in our workflow builder playground as a starting point and experiment for yourself.
If you're looking to automate this part of your operation, speak to an expert and we'll show you what it would take to set up.
The workflows at a glance
- #1 - The basic check-in. A timestamped, GPS-tagged log of what almost went wrong, rated by whoever saw it.
- #2 - With written guidance. The basic check-in plus guidance panels on what counts as a near-miss and what happens after.
- #3 - With a signature. The basic check-in plus a reporter signature at sign-off.
- #4 - With photo evidence. The basic check-in plus a photo of the exact hazard or the spot where it happened.
- #5 - With Poppi checking the photo. The basic check-in plus a photo that Poppi reviews the moment it's saved, telling the team chat if it spots a problem.
Article Content
#1 - The basic check-in
Who it's for: Single-site businesses just starting to track near-misses, with no formal reporting form yet.
What it is: A near-miss report is a short record of something that almost went wrong but did not, logged so the pattern can be spotted before it turns into a real incident. This version is four steps on a phone: type what happened, tap the time, drop a location pin, and rate how close it came on a scale of 1 to 5. Each completion is one stamped record. Whoever saw the near-miss fires the canvas there and then, and the running list of completions becomes the log.
In practice: Take a single-site garden centre. A team member is moving stacked pallets in the yard when one slides off the forklift forks and lands where a customer had been standing seconds earlier. No one is hurt. Instead of mentioning it at the next team huddle and forgetting it, they open the canvas, type "pallet slipped off the forks in the yard, near the trolley bay", tap the time, drop a pin on the yard, and rate it a 4. Submitted. The duty manager sees it that afternoon, not three weeks later when it happens again with someone standing there.
Why it works: The whole value of a near-miss is that it is free information. Nothing went wrong this time, so the only cost is the 30 seconds it takes to log it. A stamped, located, rated record means the manager can count near-misses by place and by severity. Three slips on the same corner stop being three forgotten stories and start being a pattern that gets fixed.
Steps included:
- 1 text input (what happened)
- 1 date/time input (when it happened)
- 1 location step (GPS capture)
- 1 rating scale (how close it came, 1 to 5)
#2 - With written guidance
Who it's for: Sites with rotating or shift-based staff who all need to report the same way, even on their first week.
What it is: The basic check-in plus two guidance panels woven into the canvas. The first panel explains what counts as a near-miss, with plain examples: a slip that did not end in a fall, a burn that did not quite touch the skin, a near collision in the yard. The second panel explains what happens after a report is logged, including the point that no one gets in trouble for logging one. A new starter on their first shift reports the same way a five-year veteran does, without anyone briefing them in person.
In practice: Take a busy distribution warehouse running three shifts a day. Staff rotate, agency cover comes and goes, and most people have never been told what a near-miss actually is. One picker sees a colleague step backwards off a kerb in the loading bay and catch themselves on a racking upright. Is that worth logging? The "what counts" panel answers it on the spot: yes, a slip that did not end in a fall is exactly the thing. The "what happens after" panel reassures them their colleague will not be disciplined. The report gets logged instead of shrugged off, and the duty manager sees the loading-bay kerb come up three times in a fortnight.
What it adds to the basic check-in:
- A "what counts as a near-miss" panel at the top, with concrete examples so people stop guessing
- A "what happens after you log it" panel that reassures staff and explains the next steps
- A clear no-blame message, which is what gets the quiet near-misses reported instead of buried
Why it works: The two biggest reasons a near-miss never gets logged are "I did not think that counted" and "I did not want to drop my mate in it". Written guidance sits inline at the moment of reporting and answers both. The picker reads the examples right before they type, and reads the no-blame line right before they submit. It is not a poster in the break room they walked past once. It is on the screen at the moment of the task.
Steps included:
- 1 guidance note (what counts as a near-miss, with examples)
- 1 text input (what happened)
- 1 date/time input (when it happened)
- 1 location step (GPS capture)
- 1 rating scale (how close it came, 1 to 5)
- 1 guidance note (what happens after you log it)
#3 - With a signature
Who it's for: Businesses where the reporter's name needs to be signed against the record, or sites that get audited and need a traceable trail.
What it is: The basic check-in plus a signature captured at sign-off. The four core steps stay the same; the signature confirms who reported it and when they signed it. The reporter's finger-drawn confirmation goes on the same record as the account, the time, the pin, and the rating.
In practice: A food-production site gets audited twice a year. Every near-miss report logged on the floor ends with a finger-drawn signature confirming the account is accurate. When the audit lands, the safety manager pulls 40 near-misses at random and shows the auditor a signed, time-stamped record for every one, with a closeness rating and a pin. The auditor can see the site is not just collecting reports but standing behind each one. What used to be a scramble through a paper book turns into a 20-minute review on screen.
What it adds to the basic check-in:
- A signature step at the end of every report
- A reporter confirmation on the same record as the account, timestamp, and pin
- A signed near-miss trail at the level an auditor expects to see
Why it works: A near-miss report has to be honest to be useful, and a signature is what puts the reporter's name to the account. Captured at the same moment as the report, the signature adds: and the person who saw it confirms this is accurate. It is the reporter standing behind their account.
Steps included:
- 1 text input (what happened)
- 1 date/time input (when it happened)
- 1 location step (GPS capture)
- 1 rating scale (how close it came, 1 to 5)
- 1 signature (sign-off)
#4 - With photo evidence
Who it's for: Businesses across more than one site that want photo evidence of the area or the kit that was involved.
What it is: The basic check-in plus a photo step, taken at the same moment as the location pin. The reporter takes a quick shot of the exact hazard or the spot where it happened: the frayed cable, the wet floor with no sign out, the pallet stacked too high. The photo lands in the same record as the time, the pin, and the closeness rating. A manager reviewing reports across several sites can see the actual condition, not just read about it.
In practice: Take a three-site manufacturing business with a shared health and safety lead. On the night shift at one plant, a machinist notices a machine guard hanging loose after a near-miss where a sleeve nearly caught. They log the report, rate it a 5, and take a photo of the loose guard. The next morning the safety lead opens the report from another site entirely, sees the photo, and recognises the same guard model is fitted on two other lines. One typed report would have been a line in a list. The photo turns it into an action across all three sites that same day.
What it adds to the basic check-in:
- A photo step after the rating, capturing the area or the kit involved
- Visual proof of the actual condition, which a typed description alone cannot give
- Something a manager at another site can act on without travelling to see it in person
Why it works: A description is one person's account. A photo is the thing itself. Captured at the moment, on the same device, in the same record, it cannot be staged or reconstructed later. It is the difference between "someone said a guard was loose" and "here is the loose guard".
Steps included:
- 1 text input (what happened)
- 1 date/time input (when it happened)
- 1 location step (GPS capture)
- 1 rating scale (how close it came, 1 to 5)
- 1 photo of the exact hazard or the spot where it happened
#5 - With Poppi checking the photo
Who it's for: Businesses where the photo gets taken but nobody reviews it. Multi-site groups where head office can't look at every site's reports.
What it is: A photo-checked near-miss report is the basic check-in plus a photo that Poppi (AI) reviews the moment it's saved. Poppi answers one question about the image, set by you: does the photo clearly show the hazard, or the exact spot where the near-miss happened? If the answer is no, Poppi posts what it spotted to the team chat, so the problem gets sorted before everyone moves on.
In practice: A three-site hospitality group reports near-misses across their kitchens. After a shift, a member of staff photographs the slipped blade on a food processor to log a near-miss. Poppi reads the photo: the blade is clearly visible, the angle shows the exact hazard. Verdict yes, and nothing changes. On a busy Friday evening another staff member submits a rushed photo that is too dark and blurry to see the exact spot where a glass nearly broke. Poppi answers no and posts the reason to the team chat: "The photo is too dark to see the exact spot clearly. Please retake a clearer photo of the hazard." The staff member takes a second photo while still on site, and the problem is sorted that night instead of becoming a mystery.
What it adds to the basic check-in:
- A photo of the exact hazard or the spot where it happened, checked the moment it's saved, not just stored
- A team chat message with Poppi's reason the moment a photo fails the check
- The manager stops being the only person who ever looks at near-miss photos
Why it works: The check happens in the seconds between the photo being taken and the person leaving, the only moment the problem is still cheap to fix. A manager reviewing photos the next morning can only record that the photo is no good; Poppi catching it at the time gets it retaken on the spot.
Steps included:
- 1 text input (what happened)
- 1 date/time input (when it happened)
- 1 location step (GPS capture)
- 1 rating scale (how close it came, 1 to 5)
- 1 photo of the exact hazard
- 1 Poppi decision (judges the photo against your question)
- 1 Poppi action (posts to the team chat if the photo fails the check)
How to pick the right version
You don't need to know our product to choose. Every version here is the basic check-in plus one addition, so pick the additions your business actually needs.
Do other people report near-misses?
If you report for yourself and know the process, the basic form is enough: #1. The moment rotating staff report, they need written guidance on what counts: #2.
Does the reporter's name need to be on the record?
If a manager reviewing reports later will know who reported it by context, skip this. If you need the reporter's signature captured at the moment they submit, #3 adds a signature.
Do you need to see the actual condition?
A typed description says what happened; a photo shows the hazard. If you need visual proof, #4 adds a photo of the area or kit involved.
Does anyone actually look at the photos?
If a manager genuinely reviews every photo, #4's record is enough. If photos get taken and filed unseen, #5 has Poppi (AI) check each one as it's saved, and tell the team chat when something is wrong.
Need more than one addition? Open the version with the addition that matters most in the playground and add the others as steps. That's how the product works anyway: every option here is one step added to the same four inputs.
Related workflows
- 4 ways to automate hazard spot reports
- 4 ways to automate lone worker check-ins
- 4 ways to automate maintenance fault reports
- 4 ways to automate damage reports
Conclusion
A near-miss report is a short, stamped record of something that almost went wrong, logged so the pattern gets spotted before it becomes a real incident. Every version above is the same four inputs plus one addition: guidance, a signature, a photo, or an AI check on the photo. Pick the ones your business needs and combine them in the playground.