Digital Permit to Work System

Two permits, two folders, one location

The permit to work system is the control standing between planned high-risk activity and the incidents that activity causes. On paper it depends entirely on people: an issuer who understands the location, a receiver who reads the conditions, and someone who remembers to close the permit when work stops. Each of those is a reasonable assumption individually and an unreliable one at scale, across shifts, on a site issuing forty permits a day.

The failure paper cannot address is conflict between permits. Hot work authorised on a level directly above a confined space entry; a crane lift planned over an area where another contractor holds a permit to work; an isolation applied while a permit relying on that supply is still active. Each permit was correctly assessed on its own. The hazard is in the combination, and nobody sees the combination because the two documents are in different folders held by different people. Alongside that sit the ordinary failures: permits left open after work finished, gas tests older than their validity, and closure recorded without the area being checked.

• Conflicting permits are invisible because they exist as separate documents.
• Permits stay open after work stops, so the register does not reflect reality.
• Gas test validity expires while a confined space entry continues.
• Approvals are given from an office by someone who has not seen the location.
• Closure is signed without the area being physically checked and handed back.

Solution overview

Swedish Technology builds the permit around location and time as first-class data rather than as text in a box. Each permit carries the area it applies to, its validity window, its hazard class and its conditions. That is what makes conflict detection possible: when a hot work permit is requested for a zone where a confined space entry is active, or a lift is planned over an area with live work beneath it, the system raises it to the permit authority before approval rather than after an incident.

The rest is enforcement of the disciplines paper depends on people to remember. Gas test results carry their validity period and the permit becomes non-compliant when they expire. Permits expire automatically at the end of their window and require deliberate extension rather than drifting. Closure requires the handback steps to be completed, and an open permit with no activity escalates. And because permits, site access and certification share one record, the system can check that the people performing permitted work are actually on site and hold the competencies the permit requires.

What we will not claim is that digitisation improves a weak permit process. If permits are approved by people who have not seen the location, a digital system makes that faster. The control is a competent person at the workface, and the software's job is to support that person rather than to replace them.

How the solution works

1. 1
   Request with location and hazard class The requester defines the work, its area, its window and its hazard class. Area and time are structured data rather than free text, which is what makes everything downstream possible.
2. 2
   Assess and attach controls Risk assessment, method statement, isolations, gas tests and required competencies attached to the permit rather than referenced in a folder.
3. 3
   Check for conflicts automatically The system compares the requested area and window against active and pending permits and raises conflicting simultaneous operations to the permit authority before approval.
4. 4
   Approve at the right level Routing by hazard class to the competent authority, with high-risk classes requiring a site verification step rather than a desk approval.
5. 5
   Activate and monitor The permit becomes live at the workface with its conditions visible, validity counting down, and gas test expiry enforced during the work.
6. 6
   Close with handback Closure requires the defined handback steps — isolations removed, area checked, equipment cleared — and an open permit with no activity escalates automatically.

Key capabilities

Reference architecture

Built around area and time as structured data, and around the reality that the workface frequently has no connectivity.

Deployment options: Cloud or on-premise with full offline capability on mobile devices. Permits are issued and closed in basements, plant rooms and excavations where there is no signal, so offline is a core requirement rather than a fallback.

Hardware options

Mostly software, but the devices matter because permits are issued and closed at the workface rather than at a desk.

Swedish Technology supplies and integrates site-rated devices from established manufacturers, specified for offline use in UAE site conditions.

AI capabilities

Applied to the review load and to patterns across permits that no individual sees.

• Conflict scoring — Ranks detected overlaps by severity — a hot work permit above a live confined space entry is not the same as two low-risk permits sharing a corridor — so the authority's attention goes to what matters.
• Risk assessment review — Flags method statements and assessments that omit controls normally present for that hazard class, as a prompt for a human reviewer rather than an approval decision.
• Dormant permit detection — Identifies open permits with no supporting activity — no access records, no scans — which is how the register stops filling with permits nobody closed.
• Pattern analysis — Shows which contractors, areas and shifts generate the most permit exceptions, extensions and late closures, which is a genuine leading safety indicator.

Integrations

These can be designed within project scope.

The integrations above are designed and implemented within project scope using vendor APIs, webhooks or standard connectors. They do not imply partnership, certification or endorsement by the system owner unless stated on that vendor's official pages.

Dashboards & analytics

• Live permit register — Every active permit by area, hazard class and expiry — the view needed within a minute of an incident.
• Conflicts and exceptions — Detected simultaneous operations conflicts, expired gas tests and overdue closures, ranked by severity.
• Approval performance — Time from request to approval by class and authority, which shows where the permit process is delaying work legitimately and where it is not.
• Contractor patterns — Permit volume, extensions, late closures and exceptions by contractor and area over a period.

Security & deployment

Approval authority is the security model here: permits route by hazard class to individuals authorised for that class, and authorisation is a maintained register rather than a habit. Every action — request, approval, extension, override, closure — is attributed to a named person with a timestamp, because the permit record is a safety document that may be read in an investigation. Devices operate fully offline, which means the permit issued in a basement plant room is as valid as one issued in the office, and conflict checks that could not run offline are re-evaluated on reconnection with any conflict raised immediately.

Data privacy

Permit records identify the people who requested, approved, performed and closed high-risk work, which makes them personnel records as well as safety records. That attribution is necessary — an unattributable approval is not an approval — but it means access should be scoped to HSE, operations and investigation roles rather than being generally readable.

Where the system draws on site access and certification data to verify that permit holders are present and competent, it should take the minimum needed for that check rather than a general feed of workforce movement. Under UAE Federal Decree-Law No. 45 of 2021 the purpose should be clear and proportionate, and permit records generally justify a long retention because they may be required for an investigation years later — which is a reason to retain the permit record itself, not a reason to retain everything gathered alongside it.

Industry use cases

Implementation approach

1. 1
   Permit process review Examine the existing process before digitising it. If approvals are being given by people who have not seen the location, that is the problem — and software will make it faster rather than better.
2. 2
   Area and hazard model Define the zones and hazard classes the system will reason about. Conflict detection is only as good as the spatial model beneath it.
3. 3
   Authority register Establish who may approve which class, and maintain it. This is usually less formal than sites expect when it is written down.
4. 4
   High-risk class pilot Start with hot work and confined space, where the value is highest and volumes are manageable, and run alongside paper initially.
5. 5
   Conflict tuning Adjust zone granularity and conflict rules from real permits. Too coarse produces noise; too fine misses genuine conflicts, and only real data settles it.
6. 6
   Full rollout and integration Remaining permit classes, then integration with access, workforce and maintenance systems so permits connect to who is actually present.

Why Swedish Technology

• We build conflict detection between simultaneous operations, which is the one thing a paper permit system structurally cannot do.
• Area and time are structured data rather than text in a box, which is what makes that detection possible.
• We will tell you when digitisation is not your problem — a weak approval culture gets faster, not safer.
• Devices work fully offline, because permits are issued in basements and excavations rather than at desks.
• Permit conditions are presented in the receiver's language, because a condition that is not understood is a signature rather than a control.

Limitations & prerequisites

• Digitising a weak permit process produces a fast weak process. If approvals are given without the location being seen, the software will not fix that — a competent person at the workface is the control.
• Conflict detection depends on the accuracy of the area model. Zones defined too coarsely generate noise; defined too finely they miss genuine conflicts.
• The system checks that documents exist and are current; it cannot assess whether a risk assessment is technically adequate. That remains a competent person's judgement.
• Gas test integration depends on the instruments in use. Where readings are entered manually, the system enforces validity but cannot verify the reading itself.
• Offline operation means some conflict checks are evaluated on reconnection rather than at the moment of issue, which is a real gap in fully disconnected areas.
• References to safety and labour obligations are general guidance, not legal advice; confirm requirements with the relevant authority and your client's standards.

FAQ

Related solutions

• Interactive Experiences & Smart Visitor Solutions
• Construction Site Visitor Management
• Construction Workforce Tracking
• AI PPE Detection
• Construction Equipment Tracking
• Contractor Access Management System
• Warehouse Safety AI Analytics
• RFID Asset Management & Tracking