Use case · Employee safety
Somebody opening alone at six, closing alone at eleven
This is the one use case where the requirement is not evidence. It is somebody knowing now, and being able to do something. Everything on this page is designed backwards from that, and where the honest answer is a person, not a device, we say so.
- CSLB C-7 #976737
- BSIS ACO #7501
- CMMC certified
- SAM.gov registered
- Est. 2008
The problem
The hours when one person is the whole building
Almost every commercial operation has them. The person who opens at six and is alone in an empty building for ninety minutes. The person who cashes up and locks the shutter at eleven and then walks across an unlit lot. The receptionist who is the only physical barrier between the street and everybody else, at a desk with nothing behind it. The driver doing a drop at a site with nobody on it. The warehouse supervisor doing a final walk. The letting agent showing a unit to a stranger.
What goes wrong is a short list and it is not mostly dramatic. Someone walks in behind a member of staff at opening. Someone at the counter becomes aggressive and there is nobody to call and nowhere to go. Someone who has been asked to leave will not. A confrontation in a car park at shift change. And underneath all of them, the ordinary version that costs the business the most: staff who have stopped feeling safe doing a normal part of their job, and who deal with it by changing what they do — not opening alone, not doing the walk, not taking the late shift, or leaving.
The operational cost is a staffing cost long before it is an incident cost. Two people rostered for a job that needs one, because nobody will do it alone. A shift that is hard to fill. An HR conversation. An insurer or an auditor asking what provision exists for lone working. And a manager carrying the whole thing personally, because they are the one who gets the call.
It resists solving because it is the problem furthest from what a security system is naturally good at. A camera produces a record. A recorder produces a record. An alarm produces a signal about a building. None of those is what a person standing at a counter with somebody shouting at them actually needs, which is for somebody else to know, right now, without it being obvious that they told them. That is a narrow requirement, it is met by a small number of specific things, and it is frequently missing from sites that have spent a great deal on everything else.
The data
What San Diego records, and what nobody records
There is no workplace-violence statistic on this page, because we hold no primary source for one — and this is the last subject on which an unsourced number should be published. What follows is what is genuinely countable.
| Year | Robbery offences |
|---|---|
| 2015 | 1,378 |
| 2020 | 1,207 |
| 2023 | 1,205 |
| 2024 | 1,069 |
| 2025 | 819 |
Source: Federal Bureau of Investigation, Crime Data Explorer, agency ORI CA0371100 for San Diego Police Department — cde.ucr.cjis.gov/LATEST/webapp. Period: selected years from the 2015–2025 series. Type: COUNT of offences reported by the agency. Retrieved: 2 September 2026. The 2025 rate was 58.1 per 100,000 residents served. UCR participation is voluntary. Robbery is falling in San Diego — 1,069 offences in 2024 against 819 in 2025 — and, nationally, the FBI reports violent crime down and 47.4% of it cleared in 2025. These are city-wide totals covering every setting, not commercial-premises figures: no primary source separates a robbery at a business from one on a street, so this cannot be read as a count of incidents at workplaces.
| Call type | Calls dispatched, 2025 |
|---|---|
| Burglary alarm | 17,075 |
| Robbery alarm | 971 |
| Panic alarm | 918 |
| Vehicle alarm or audible | 305 |
Source: San Diego Police Department, Police Calls for Service, City of San Diego Open Data Portal — data.sandiego.gov/datasets/police-calls-for-service. Period: calendar year 2025. Type: COUNT of calls dispatched, computed on de-duplicated rows for this fact pack. Retrieved: 2 September 2026. These are calls dispatched, not activations — anything resolved before dispatch, or resolved by a monitoring centre, never appears in this data. The dataset also excludes several sensitive incident types by design. Robbery and panic alarms together account for fewer than 1,900 dispatched calls against 17,075 burglary alarms, which tells you these are a small and distinct category, not how often duress is raised anywhere. The full analysis of this dataset, with the method and every caveat, is on after-hours coverage.
What we are not going to give you is a workplace-violence percentage. We hold no primary source for one, and the figures circulating on this subject come from vendors and content sites. On a page about people's physical safety, an unsourced number is worse than no number. The argument for duress provision does not rest on a statistic. It rests on the fact that the person at the counter has no other way to tell anybody, and that the alternative arrangement — a siren, or a phone in a pocket that has to be visibly used — is not one.
50%
ESTIMATEFrom a voluntary, self-selected survey of 66 retail companies representing 143 brands. Not a census and not government data. Quoted because 'the same people, repeatedly' is the pattern staff describe most often — and the same study reports shoplifting incidents DOWN 12.4% in 2025, so the concern is not driven by a rising volume of incidents. It is retail-specific and does not describe other sectors on this page.
| Contributing factor cited | Retailers |
|---|---|
| Repeat offenders | 50% |
| Organised retail crime incidents | 40% |
| Walkout or pushout theft | 37% |
National Retail Federation, Impact of Theft & Violence 2026, covering fiscal 2025, released 30 July 2026. Type: ESTIMATE from a voluntary survey. Retrieved: 2 September 2026.
Source for the figures above: National Retail Federation, Impact of Theft & Violence 2026, covering fiscal 2025 \u2014 nrf.com, released 30 July 2026. Type: ESTIMATE from a voluntary, self-selected survey of 66 retail companies representing 143 brands; not a census and not government data. Retrieved: 2 September 2026.
Two lines we hold
Egress, and fire
A page about keeping staff safe that also sells door control has an obligation to be explicit about the two places where our equipment must not be load-bearing, so here they are before anything else on this page is discussed.
Nobody's way out of a building may depend on our system. A person leaving in an emergency must not depend on an access-control system being powered, being online, or recognising them. Egress arrangements and fire doors are governed by the building code and by your fire and life-safety disciplines, and they set what any access system is permitted to do at those openings. We design inside those constraints, and where a specification handed to us would make egress depend on our equipment, we will decline it and say why instead of fitting it and leaving the problem on site.
We do not hold a C-10 or a C-16, so we do not work on fire alarm systems. That is coordinated with your own licensed fire alarm contractor — we do not perform or monitor it. Anything that interfaces door hardware to a fire alarm is their determination and the building's, not ours, and we would rather tell you that at the survey than let it be discovered at an inspection. Our monitoring scope is intrusion, video and access only.
There is a third line and it belongs alongside these two. Where the answer to your problem is a person — somebody present at shift change, somebody to stand between a member of staff and a person who will not leave, somebody to remove them — that is guard and patrol work, licensed separately in California under a Private Patrol Operator licence. We coordinate it through Guardian Integrated Security, Inc., of which we are a DBA, and we do not present ourselves as a guard company. This is the use case where that answer is most often the right one, and selling equipment instead would not fix it.
What reduces it
Six things that change the outcome
Ordered by how directly each addresses the actual requirement, which is that somebody else knows now.
Duress that reaches a person
A monitored hold-up or duress signal, raised without it being obvious, arriving as a distinct signal type instead of a generic intrusion and triggering a defined procedure — on a device the person actually has on them. That is commercial alarm monitoring.
A boundary somebody has to be let through
Controlled doors behind reception, on the back of house and on the cash office, so a person who has walked in has not automatically reached everybody — designed so that egress never depends on it. That is commercial access control.
Deciding before the door opens
A door station that lets somebody see and speak to a caller while there is still a locked door between them moves the decision earlier, which is most of the benefit at a reception or a lone-working site. That is video intercom and door entry.
Somebody watching at the hours that matter
Opening, close and shift change are narrow, predictable windows. Watched coverage during those windows, with an operator able to speak into the space, is the closest thing to a colleague being present. That is remote video monitoring.
Coverage of the routes, not the building
The walk from the door to the car, the stair core, the rear exit and the yard — positions specified to identify at night, because that is where staff feel exposed and where a great deal of this happens. That is commercial security camera installation.
Pathway to the awkward positions
Duress devices, door stations and outdoor cameras on a route all need something behind them, and whether a position can be reached and powered decides the design before any device is chosen. That is structured cabling.
Who has it worst
Where lone working is structural
Any business with early opens, late closes or a public-facing counter has this. Three types have it as a standing condition of how they operate.
- Retail. Opening and closing alone, a counter facing the street, cash handling, and the pattern staff describe most often — the same individuals returning, which turns an incident into an ongoing situation. That is retail security and loss prevention.
- Managed office and multi-tenant property. Reception and concierge staff who are the only physical boundary in the building, out-of-hours contractors, and parking structures where tenants and staff walk alone at either end of the day. That is office and property management security.
- Apartment communities and managed residential property. On-site managers and maintenance staff dealing with residents and visitors alone, in stairwells, plant rooms and parking areas, where the coverage question is bounded by real privacy expectations. That is multifamily and apartment security.
The other verticals are on all industries and the full service list is on all commercial services. The related problems are parking lot incidents for the walk to the car, and trespassing and loitering where the same people keep returning to the site.
Everything on this page is designed for commercial premises. We do not work on private residences, though managed residential property is in scope. And nothing here is a promise about how quickly anybody responds to anything — there is no response-time claim anywhere on this site, and on a page about duress that omission is deliberate, not accidental.
At survey
What we would look at on your site
This survey is mostly a conversation with the people who do the shifts — not a walk round with a manager. It produces the shortest equipment list of any use case in this section.
Who is alone, when, and doing what
Named tasks and named hours: opening, cashing up, the final walk, the late delivery, the weekend visit. Written down, because the list is always longer than the first answer and usually includes somebody nobody had counted.
What has already happened
The incidents, the near misses and the things staff have quietly stopped doing. The last of those three is the most informative and the one least likely to be in any log.
Where a person can go, and how far they get
From the entrance, from the counter, from a delivery door. Whether there is any controlled boundary behind the public face of the building, and how quickly somebody who has walked in reaches the back of house.
The routes at the hours they are used
The walk from the door to the car at close, the stair core, the rear exit, the yard. Lit or not, overlooked or not, and what any camera there would actually produce at that hour.
Egress and any fire interface, identified early
Which openings are on an escape route, what is already interfaced to the fire system, and therefore what door control is permitted to do at each. Established before design, and coordinated with your fire contractor rather than assumed by us.
What happens after the button is pressed
The procedure, the call list, who is actually reachable at that hour and what they are expected to do. A duress device attached to no plan is a device, and this part costs nothing to get right.
Findings go in writing whether or not you go on to contract with us. On this use case in particular the report will say plainly where we think the answer is a procedure, a rota change or an officer — not equipment — and that is frequently the larger half of it.
Common questions
Employee safety questions
Three things, and none of them is a camera. First, a way to raise an alarm that reaches a person, not a siren — a monitored duress or hold-up signal, on a device they actually have on them instead of one screwed under a desk they may not be standing at. Second, a way to see the approach before committing to it: the ability to look at the car park and the entrance from a phone before getting out of the car changes the risk of opening alone more than anything installed at the door does. Third, a routine that somebody else is part of, so that a missed check-in is noticed. The technical parts are ordinary and we can fit them. The routine is yours, and a system without one is a device nobody uses.
RelatedPanic and duress buttons
A monitored signal path to a person, which is the whole point of it — a device that only makes a noise in the room where the incident is happening is not a duress alarm, it is a noise. A hold-up or duress signal is designed to be raised without being obvious, to arrive at a monitoring centre as a distinct signal type, not a generic intrusion, and to trigger a defined procedure. It is also a signal type that authorities treat differently from an intrusion alarm, which is exactly why it must not be used casually or tested without arrangement. We will not tell you how quickly any particular response arrives, because that is not ours to promise and no such promise appears anywhere on this site.
RelatedPanic and duress buttons
Where the person will be when they need it, which is very often not where it gets installed. Fixed buttons belong under a counter within reach of the position somebody actually stands in, at a cash office desk, and at a reception station — not somewhere that requires walking. But the fixed button is the weaker half: a wearable or mobile credential goes wherever the person goes, including the stockroom, the yard and the car park, which is where a good share of incidents actually occur. Where it should not go is anywhere it can be pressed accidentally by a knee, a chair or a delivery, because a duress signal that goes off three times a month stops being taken seriously and that is a genuinely dangerous outcome.
RelatedPanic and duress buttons
It can, and mostly by moving the decision earlier. The value of a door station is that somebody decides whether to admit a person while there is still a locked door between them instead of after the person is already inside and standing at the desk. For a reception, a back office, a cash office or a site where staff work alone, that changes the shape of the interaction entirely. It works best where there is a real boundary to control — a lobby, an airlock, a counter with a controlled door behind it — and is much weaker where the space is open to the street by design. What it is not is a substitute for a plan about what staff do when somebody they have declined refuses to leave.
RelatedPanic and duress buttons
It must not, and this is the most important sentence on this page. A person leaving a building in an emergency must never depend on an access-control system working, being powered, or recognising them. Egress arrangements, fire doors and anything on a fire alarm interface sit with the building code and with your fire and life-safety disciplines, and they govern what any access system is allowed to do at those openings. We do not hold a C-10 or a C-16, so we do not work on fire alarm systems; that is coordinated with your own licensed fire alarm contractor, not performed by us. We design door control to sit inside those constraints, and if a proposal we are given would make egress depend on our equipment, we will say no and explain why.
RelatedPanic and duress buttons
Our default is no, and we will not change it on request alone. California treats the intentional recording of a confidential communication without the consent of all parties as prohibited under Penal Code section 632. The statute defines a confidential communication as one carried on in circumstances that may reasonably indicate a party desires it to be confined to the parties, and expressly excludes circumstances in which the parties may reasonably expect it may be overheard or recorded — so the rule and the exception have to be read together, and whether a particular reception, lobby or counter falls inside or outside the definition depends on the facts of that space. That is not a determination we make for you. This is not legal advice: take it to your counsel first, and we will design to whatever they tell you.
RelatedPanic and duress buttons
Partly, and being honest about which part is useful. What a system can address is the environment: light on the route, coverage of the route itself and not just the building it leaves, and the ability for somebody to be watching during those specific twenty minutes at the end of a shift. What a system cannot address is the feeling of being alone in a large dark space, and on a good number of sites the intervention that actually works is a change to the routine — leaving together, a different door, a reserved area closer to the building, or a member of staff or an officer present at shift change. We would rather say that than sell a camera that makes a recording of something nobody wanted to happen.
RelatedPanic and duress buttons
Technically, quite a lot: coverage that produces a usable identification, not just a shape, a record with accurate times that supports whatever route you and your advisers decide to take, controlled entry so the person cannot simply walk into the back of house, and duress provision for the staff who deal with them. What we cannot do is any part of the response involving a person — removing somebody, standing between them and your staff, or serving anything on them. That is guard and patrol work, licensed separately in California under a Private Patrol Operator licence, and where a site genuinely needs officers we coordinate through Guardian Integrated Security, Inc., of which we are a DBA. This is the situation where we most often tell a buyer that equipment is the smaller half of the answer.
RelatedPanic and duress buttons
We cannot put a person on your site. Officers are Private Patrol Operator work, and that licence — PPO #121089 — is Guardian Integrated Security, Inc.'s. We are a DBA of Guardian, so it is the same company's other licence, arranged through them and never sold from this page. If a site is under active threat, that call should go to Guardian or to the police before it goes to a camera contractor. The systems themselves are ours: CSLB #976737 covers installing the duress buttons, cameras and door hardware, and BSIS ACO #7501 covers monitoring the signals they raise.
RelatedPanic and duress buttons