How Torchpass scores a building
Every formula, weight and answer value behind the three scores. Version 1.1.0, published November 1, 2026.
Torchpass gives three scores: knowledge risk, asset list health (or contact list health on the front-of-house path) and AI readiness. This page sets out exactly how each one is calculated, what each answer is worth, and where the method is still being tested. Nothing here is hidden from the results you see.
Principles
- Answers stay on your device. All scoring runs in your browser.
- Simple enough to check by hand. Every score is a short formula you can reproduce with a calculator.
- Life safety counts most. Areas where a failure could hurt people or shut the building carry more weight.
- A first look. Scores come from self-reported answers. A walk-through confirms the details.
Three paths
The first question asks about your role. Each path scores the areas that person is most likely to know, using the same formulas.
| Path | Who it's for | Areas scored | List scored |
|---|---|---|---|
| Front of house | Reception, office and workplace teams | 8: reception and visitors, client and VIP visits, meeting rooms and AV, pantry and catering, mail and packages, plus three critical areas: emergency roles and evacuation, first aid and AED, building access and badges | Vendor and contact list |
| Facilities | Engineers, technicians, coordinators and facility managers | 9: building and site information, HVAC plant, electrical, fire and life safety, health and safety, access control, building automation, plumbing and water, and front of house | Equipment (asset) list |
| A bit of both | Office managers and small-site leads | 13: the eight facility areas plus the five non-critical front-of-house areas. Health and safety and access control already cover the critical front-of-house topics. | Equipment (asset) list |
Because each path scores different areas, compare results within the same path. Results show which path was used.
1. Knowledge risk score
Each area of the building (HVAC plant, electrical, fire and life safety and so on) gets its own score from four factors that multiply together:
| Factor | Answer | Value |
|---|---|---|
| Criticality | Critical (life safety or shutdown risk) | 3 |
| Core operations | 2 | |
| Concentration Who could run it alone | Nobody on staff | 1.00 |
| 1 person | 0.85 | |
| 2 people | 0.50 | |
| 3 or more people | 0.20 | |
| Turnover When the main role may change | Within 6 months, or announced | 1.00 |
| Likely within 12 months | 0.70 | |
| Possible in 1 to 3 years | 0.50 | |
| Stable for 3+ years | 0.30 |
Documentation gap
Four checks: written procedures exist, procedures were updated in the last 12 months, asset data is complete, and an expert has reviewed the documents. The share of checks in place becomes a documentation score from 0 to 1:
So nothing documented gives a gap of 1.0, and everything documented gives 0.1. In detailed mode, each area also has an item-by-item checklist (Yes = 1, Partly = 0.5, No = 0, Not relevant = left out). The documentation score is then half the four checks and half the checklist.
Building score
The building's knowledge risk score is the average of its area scores, weighted by criticality, so critical areas count 1.5 times as much as core ones. If any critical area scores 40 or higher, the building is rated high risk whatever the average. Areas marked not relevant are left out. Areas you add yourself are scored the same way.
Bands: 0–14 low · 15–39 moderate · 40 and above high.
2. Asset list health
| Part | Weight (quick) | Weight (detailed) |
|---|---|---|
| Coverage: how much equipment is listed | 45% | 35% |
| Details recorded (core fields count double) | 35% | 25% |
| How recently the list was checked against the building | 20% | 15% |
| Process for updating the list when equipment changes | – | 10% |
| Physical tags that match the list | – | 8% |
| Organized by site, building, floor and system | – | 7% |
The core record is location, make and model, serial number and install date. Answers such as “Not sure” score low, because an unknown list behaves like an incomplete one.
Contact list health (front-of-house path)
| Part | Weight (quick) | Weight (detailed) |
|---|---|---|
| Whether a new person could find who to call | 45% | 35% |
| Details recorded (core fields count double) | 35% | 30% |
| How recently the contacts were checked | 20% | 15% |
| Where the list lives (one shared place, scattered, or one person's phone) | – | 20% |
The core record is the company and what they do, a named contact, phone and email, and an account or contract number.
3. AI readiness
Five pillars, each scored from 0 to 100, combined as a weighted average:
| Pillar | Weight | Built from |
|---|---|---|
| Data foundation | 30% | Where equipment data lives (35%), asset list health (40%), areas with complete asset data (25%) |
| Work history and naming | 20% | How work orders are recorded, plus naming standards in detailed mode |
| Documented know-how | 20% | Areas with procedures (30%), current procedures (30%), reviewed procedures (20%), knowledge held by 2+ people (20%) |
| Connected systems | 15% | Access to BAS and meter trend data |
| Ownership and rules | 15% | A named data owner and rules for using AI (detailed mode; quick mode re-weights the other four) |
On the front-of-house path, the same pillars and weights apply with front-of-house inputs: the information foundation uses contact list health and where procedures and vendor details are kept; request history replaces work orders; and connected systems means data from visitor and room booking systems. The use cases are a front desk assistant, request triage, room and space insights, and visitor and event automation.
Foundation cap: if the data foundation scores under 40, AI readiness is capped at 44, because every AI use depends on it.
Bands: under 45 not ready · 45–69 partly ready · 70 and above ready.
AI use cases
| Use | What it needs |
|---|---|
| Technician assistant | Documented know-how 60+, data foundation 45+ |
| Work order triage and reporting | Work history 55+, data foundation 50+, a CMMS |
| Fault detection and energy savings | Connected systems 65+, data foundation 50+ |
| Predictive maintenance | Work history 75+, connected systems 65+, data foundation 70+ |
A use is marked ready when every threshold is met, close when each gap is 15 points or less, and not yet otherwise.
Informed by recognized practice
The areas and checklist items draw on widely used facilities standards and regulations. Torchpass is not certified, endorsed or reviewed by any of these bodies, and a screening is not a compliance audit.
| Area | Related references |
|---|---|
| Overall approach | ISO 41001 (facility management systems) |
| Asset list and building information | ISO 55000 and ISO 55001 (asset management) |
| HVAC plant | ASHRAE Standard 180 (inspection and maintenance of commercial building HVAC systems) |
| Electrical and emergency power | NFPA 70B (electrical equipment maintenance), NFPA 110 (emergency and standby power) |
| Fire and life safety | NFPA 72 (fire alarm), NFPA 25 (water-based fire protection inspection and testing) |
| Health, safety and security | OSHA 29 CFR 1910.147 (lockout/tagout), 1910.1200 (hazard communication), 1910.38 (emergency action plans) |
| Plumbing and water | ASHRAE Standard 188 (Legionellosis: Risk Management for Building Water Systems) |
| Emergency roles and evacuation (front of house) | OSHA 29 CFR 1910.38 (emergency action plans) |
| First aid and AED (front of house) | OSHA 29 CFR 1910.151 (medical services and first aid) |
OSHA rules apply to employers, and codes such as NFPA standards apply where your local authority has adopted them. The references above show where each area's checklist comes from. They are not a list of the rules that apply to your building.
Background research
These studies shaped why Torchpass looks at people, records and AI readiness together. They are context only and are not used in any score.
- Nearly 40% of facility managers expect to retire within ten years, and filling an FM vacancy now takes 17 weeks on average. IFMA, 2026 Global Salary Report
- 67% of US facilities teams already use AI, and data quality or integration is the most-cited barrier to expanding it. Johnson Controls, 2026 (260 US facilities managers)
- 52% of UK facilities professionals lack confidence that their data is accurate enough to support AI. MRI Software, 2026 (188 UK respondents)
- Workers across industries say 42% of the knowledge in their role is unique to them. Panopto, 2018, as reported by HR Dive (1,000 respondents)
These figures are checked once a year and updated when newer research is published.
Limitations
- Scores reflect the answers given. Optimistic answers produce optimistic scores.
- The weights are a reasoned starting point, not yet calibrated against outcomes. Treat a 72 and a 78 as about the same.
- Torchpass does not assess equipment condition, code compliance or safety. Those need qualified people on site.
How the method improves
Every results page asks whether the score matches what you know about your building. That feedback, along with reviews from facility managers, is used to tune the weights. Changes are versioned and listed below, so earlier results can always be explained.
Changelog
- v1.1.0 · November 1, 2026. Added three paths: front of house, facilities, and a bit of both. New front-of-house areas, contact list health, and front-of-house versions of the AI readiness questions and use cases. Facilities results are unchanged, and screenings started before this version continue on the facilities path.
- v1.0.1 · October 31, 2026. Wording only, no change to any score: matched the official titles of ASHRAE Standards 180 and 188, and clarified when OSHA rules and NFPA codes apply.
- v1.0 · October 31, 2026. First published method: knowledge risk across nine areas with custom areas, asset list health, five-pillar AI readiness with use-case thresholds, and quick and detailed modes.