Case study · Building access · Raspberry PLC + AI
Only authorised faces open the door. A Raspberry PLC controls who enters each floor.
How an office building uses a presence sensor, an AI camera and a Raspberry Pi based PLC as the central controller to let only authorised people into the building and onto their floors.
Only authorised people should reach each floor.
Without ID cards, fingerprints or extra guards.
An office building implemented an access control system for its employees. The aim is to make sure only authorised employees and people can access the building and their respective floors, and to keep unauthorised individuals out.
The challenge
Recognise each person at the entrance and at the door of every floor.
The system has to capture the face of whoever approaches, decide whether that person may enter that specific building or floor, open the door or raise an alarm, and keep a record of every access.
Security
Protection against theft and intruders, while employee privacy is kept: no personal information such as fingerprints is captured.
Access records
Entry and exit times are recorded, which makes it easier to manage timetables and check that they are kept.
Granular access
Each employee can move only through the areas allowed by their access level, set according to role and responsibility.
Lower operating costs
No additional security guards are needed and ID cards are avoided, which reduces the cost of the security system.
How it works
From face detection to door opening.
The presence sensor, the AI camera and the recognition run around the Raspberry PLC. Choose who approaches the door and whether the building is online to see the path the decision takes.
Presence sensor and AI camera
The presence sensor included in the camera detects a person approaching the entrance of the building or the access door of a floor. The AI camera then captures an image of the face and sends it to the system for analysis.
- Input: presence sensor in the camera
- Capture: face image from the AI camera
- Central controller: Raspberry PLC
Illustrative simulation. It does not use real images or data from the installation.
Open the door or raise the alarm
The system compares the image with a database of authorised faces. If the person is authorised for that building or floor, the corresponding door opening mechanism is activated. If not, an alarm is triggered and an alert is sent to the security officers.
- Comparison with the set of authorised people
- Authorised: door opening mechanism
- Not authorised: alarm and alert to security
Schematic illustration.
Storage server and access reports
A storage server keeps the access information of each employee: identification, name, job title and permissions. Accesses are tracked to generate reports, and if someone enters with unusual objects such as parcels or bags, the event is recorded for further analysis.
- Server: identification, name, job title, permissions
- Reports from the tracked accesses
- Unusual objects recorded for analysis
Access log
Illustrative simulation. It does not use real images or data from the installation.
System architecture
Every request, end to end.
The operating diagram of the project, redrawn. The camera captures the subject, the Raspberry PLC runs the API module, and the API talks to the facial recognition, the database and the recognition system.
| Flow | Between | Type | Purpose |
|---|---|---|---|
| Image capture | Subject → camera | Presence sensor and AI camera | Captures the face of the person approaching |
| Run | Raspberry PLC → API module | Program | The PLC runs the API that coordinates the process |
| Request and answer | API ↔ facial recognition | Socket | Analyses the captured face |
| Request and answer | API ↔ recognition system | Socket | Compares the visitor with the authorised people stored on the server |
| Confirm, deny, authorise | API ↔ database | Internal database | Keeps validating when there is no internet connection |
| Door opening | System → door | Opening mechanism | Lets an authorised person in |
| Alarm | System → security | Alert | Warns the security officers when the person is not authorised |
Why Industrial Shields
Industrial hardware, open solution.
Open solution
No licence fees for the software that runs the system.
Modular
The system can be expanded in the future, for example with more doors or floors.
Built for industrial use
Equipment designed and manufactured for industrial use, at affordable prices.
Hardware used
One industrial Raspberry Pi as the central controller.
Raspberry PLC
An industrial PLC based on Raspberry Pi acts as the central controller: it receives the capture from the camera and runs the API module that coordinates recognition, database and server.
- Role
- Central controller, runs the API module
- Communication
- Sockets with the client server
- Offline mode
- Internal database
- Base
- Raspberry Pi with Linux
The case does not publish the exact Raspberry PLC model. The image shows the Raspberry PLC 21 FLEX.
Questions about this case
How does the system detect a person?
With the presence sensor included in the camera. When it detects someone approaching the entrance or the door of a floor, the AI camera captures an image of the face.
How does it decide who can enter?
It compares the captured face with a database of authorised faces and checks whether that person has permission for that building or floor.
What happens if the person is not authorised?
An alarm is triggered and an alert is sent to the security officers for further action.
Does it work without internet?
Yes. An internal database allows the same validation when there is no internet connection, and it also records people who tried to be validated but were not recognised.
What role does the Raspberry PLC play?
It is the central industrial controller: it receives the capture from the camera and runs the API module, which uses sockets to send the results and compare them with a client server.
Are fingerprints or ID cards needed?
No. Access is based on facial recognition, so no fingerprints are captured and ID cards are not needed.
Download the case study
4-page PDF with the challenge, the logic, the architecture and the hardware. No form required.
Similar project?
Tell us your doors and your rules. An engineer will answer.
Access control, surveillance or any system that combines vision and an industrial controller: describe what you need to detect, decide and record, and we will tell you which controller fits.
- Answer from our technical team
- Hardware sizing for your project
- Arduino, ESP32 or Raspberry Pi based PLCs