DEVB TC(W) No. 1/2026, Annex A
Co-supervision Platform functions, and the site data behind each one
Annex A of DEVB TC(W) No. 1/2026 lists four technologies behind SmartEye, the Co-supervision Platform, and nine functions, from the time series point cloud viewer to the AI chatbot. Every function runs on site data. Kodifly supplies that data with fixed-point and fixed-route 4D LiDAR, cameras, sensors and edge units, installed as one project.
What the platform does, in one paragraph
SmartEye is the Co-supervision Platform (CSP) that DEVB developed in-house (paragraph 6). It gives works departments, consultants and contractors one place to check workmanship, productivity, positions and dimensions off site, round the clock (paragraph 7). Annex A, paragraph 1, says pilots have run in selected public works projects since late 2024.
Annex A explains the platform in three layers. Paragraph 2 names four technologies, 4D LiDAR, edge computing, 5G fusion bonding, and Generative AI with a Large Language Model. Paragraph 3 lists the other devices it can read, cameras, sensing equipment and works-specific devices. Paragraph 5 lists nine functions for quality checks, site measurements, construction monitoring and contract administration.
A dashboard function is only as good as the data reaching it. A volume figure needs a dimensionally sound point cloud. A rebar check needs a view that sees the bars before the concrete goes in. This guide takes each item in Annex A, states the site data it depends on, and shows how Kodifly supplies it.
The four technologies in Annex A, paragraph 2
What each technology does in the platform, and what has to be on site for it to work.
| Technology | What Annex A says | What the site must supply | How Kodifly supplies it |
|---|---|---|---|
| 4D LiDAR, Video of Point Cloud | Combines time-series imaging with LiDAR point clouds for dimensional and positional information. Produced by a fixed-point or fixed-route approach (para 2(i)). | Repeated scans of the same works area, so changes can be seen over time. | ScanStation K4DL-1000 on poles or tower cranes, plus fixed-route drone LiDAR for areas a fixed unit cannot see. |
| Edge computing | Optimises point cloud size, adds colour and georeferencing, and gives zero-lux vision for round-the-clock monitoring (para 2(ii)). | Processing at the device, close to the sensor, before the data leaves site. | Kodifly edge units installed with the LiDAR. ScanStation outputs real-time colourised point clouds with centimetre-level georeferencing. |
| 5G fusion bonding | Combines bandwidth from several mobile operators to raise uplink bandwidth up to about three times (para 2(iii)). | A stable uplink from each capture point to the platform. | Kodifly plans the uplink for every capture point as part of the installation. |
| Generative AI and LLM | Turns site data into actionable intelligence through an AI-driven interactive dashboard (para 2(iv)). | Dated, located, structured records the model can read and compare. | Every scan and reading is dated and georeferenced. SpatialSense drafts site reports from the same record. |
The nine functions in Annex A, paragraph 5
The functions DEVB has developed so far. New modules, such as NEC contract administration, are being added in stages.
| Function | Data it needs from site | How Kodifly supplies it |
|---|---|---|
| (i) Time series point cloud viewer for 4D LiDAR videos | Point clouds of the same area captured again and again, aligned to one coordinate frame. | Fixed-point ScanStation watching one area over time, with fixed-route drone passes added on the same frame. |
| (ii) Dimension and volume measurement | Georeferenced point clouds with known accuracy over the stockpile, excavation or structure. | ScanStation range accuracy of ±2 cm and centimetre-level georeferencing. |
| (iii) Slope angles and ground profile measurement | Repeated scans of the slope or formation surface from a stable viewpoint. | Pole-mounted units facing the slope, with drone routes over wider formation areas. |
| (iv) Check on reinforcement bar fixing | A clear, dense view of the bars before concreting. | Units placed on tower cranes or poles with a line of sight to the deck, timed to the pour programme. |
| (v) Productivity check for MiC installation | A time series over the lifting and placing zone for each module. | Crane-mounted and pole-mounted units covering the installation area. |
| (vi) Efficiency check for tower crane operation | Continuous capture of the crane, its loads and the set-down areas. | Units on the crane or on poles with a view of its working radius. |
| (vii) Alerts system | Continuous data from LiDAR, cameras and sensors, with thresholds set for the works. | Round-the-clock capture, AI cameras and IoT sensors, delivered on the same platform. |
| (viii) Site report generation | Dated records of progress, measurements and events. | Dated scans and readings, with SpatialSense drafting site reports. |
| (ix) AI chatbot | Structured site data tagged by date and location. | Every capture is stored with its date and position, so questions can be answered from the record. |
The devices in Annex A, paragraph 3, delivered.
Apart from 4D LiDAR, the platform reads data from cameras, sensors and works-specific devices. The list is not exhaustive.
All supplied, installed and supported by Kodifly
Fixed-point ScanStation K4DL-1000
Fixed-route drone LiDAR
AI cameras
Body-worn cameras
Temperature, humidity and vibration sensors
Edge units at each capture point
Site report drafting in SpatialSense
Link to the DWSS hub
Co-supervision Platform functions, answered
How many functions does the Co-supervision Platform have?
Annex A, paragraph 5, of DEVB TC(W) No. 1/2026 lists nine functions, a time series point cloud viewer, dimension and volume measurement, slope angle and ground profile measurement, reinforcement bar fixing check, MiC installation productivity check, tower crane efficiency check, an alerts system, site report generation and an AI chatbot. New functions, such as NEC contract administration modules, are being introduced in stages.
What is the Video of Point Cloud?
It is SmartEye's 4D LiDAR output, combining time-series imaging with LiDAR point clouds to give dimensional and positional information over time (Annex A, paragraph 2(i)). It is produced by LiDAR on fixed poles, tower cranes or non-movable objects, or by small unmanned aircraft or robots on a fixed route. Kodifly supplies both, with ScanStation K4DL-1000 and fixed-route drone LiDAR.
What does edge computing do in SmartEye?
Annex A, paragraph 2(ii), says edge computing optimises the size of point cloud images, adds colour and georeferencing, and provides zero-lux vision for round-the-clock monitoring. Kodifly installs edge units with its LiDAR, and ScanStation outputs real-time colourised point clouds with centimetre-level georeferencing.
What is 5G fusion bonding?
It combines data bandwidth from multiple mobile operators into one large channel, increasing uplink bandwidth up to about three times (Annex A, paragraph 2(iii)). It matters because point clouds are large and must reach the platform from site continuously.
Which sensors can feed the platform besides LiDAR?
Annex A, paragraph 3, lists AI cameras and body-worn cameras, IoT sensors for temperature, humidity, vibration and other environmental parameters, and works-specific devices such as thermal sensors, fibre optic sensors and handheld laser scanners. The list is not exhaustive. Kodifly supplies and installs these alongside the LiDAR.
Does the platform connect to the DWSS?
Yes. The CSP should support data connection with the Smart Construction Management hub of the DWSS through an API or other means (Annex A, paragraph 4, and paragraph 8 of the circular). DWSS requirements on acceptance, handover, technical audit and the iCWP interface also apply to the CSP (paragraph 16).
Who developed SmartEye?
DEVB developed SmartEye in-house to establish the Co-supervision Platform (paragraph 6 and Annex A, paragraph 1). Kodifly supplies the site capture, hardware, software and services the platform uses on a contract.
Related guides
Fixed-point vs fixed-route LiDAR
Poles and tower cranes against drones and robots, and when to use each.
Plan the capture for each function
Tell us the site, works type and the functions in your CSP scope. We will plan the fixed-point and fixed-route capture they need.