US 
Construction Industry Deploys Rugged Jetson AI Computers & LiDAR Sensors
Rugged Jetson computers with LiDAR give autonomous construction vehicles computer vision in dust, shock, and extreme temperatures.
Construction sites are among the hardest places to deploy modern electronics: dust, mud, rain, high-pressure wash-downs, temperature swings and constant shock and vibration. A leading manufacturer of construction vehicles asked us to help it bring NVIDIA Jetson AI out onto its machines, from graders, wheel loaders and track loaders to off-highway trucks, mining trucks and dozers.
We configured sealed, fanless Jetson AI computers from Syslogic, built to work on the outside of the vehicle with no protective cabinet. They act as each machine’s vision system, processing LiDAR and other sensor data in real time. It is the kind of system we build for autonomous and unmanned systems.
Specializing in the deployment of autonomous vehicle PCs, we configured ruggedized Jetson compute platforms for the rigours of outdoor environments, wide operating temperatures, variable power sources, shock and vibration.
What the Jetson AI computers do on each machine
- 01Object recognition
- Identifying people, vehicles and obstacles around the machine as it works.
- 02Environmental perception
- Building a live 3D picture of the site from LiDAR and other sensors.
- 03Intelligent control
- Feeding that picture into the vehicle’s control, so it can react to what it sees.
- 04Real-time sensor processing
- Handling several sensor inputs at once and evaluating them simultaneously, on the machine.
- 05Three Jetson modules
- Built on Jetson TX2 NX, Xavier NX and AGX Xavier, matching AI performance to each vehicle’s sensor load.
These were earlier Jetson generations. For new projects, our Jetson module comparison sets Orin Nano, Orin NX, AGX Orin and Thor side by side on AI performance and power, and our construction AI FAQ covers where the technology is used on site today.
How LiDAR spots danger before damage
The customer takes proven automotive LiDAR into construction, to detect deviations and danger early, before serious damage occurs.
-
Scan
Thousands of pointsLiDAR sensors scan the scene continuously, taking thousands of distance measurements to build a 3D model.
-
Model
Volumes and shapesUnlike video, the 3D model can be analyzed for volumes, shapes and distances, and objects can be filtered by their volume.
-
Compare
Gain or lossA volume increase in a static scene means something has entered it, such as a person, a vehicle or spillage. A volume loss can signal a collapse in a rock face or infrastructure.
-
Act
Rules per areaDifferent rules apply to different areas, such as vehicle roads and safe walking paths, so the system flags only the changes that matter.
More LiDAR, one cable each
Each computer has two Gigabit Ethernet ports as standard. To connect the extra LiDAR sensors on the vehicles, we fitted PoE add-on boards with four more Gigabit Power over Ethernet ports on M12 connectors. The board is powered internally from the computer, so no additional power supply is needed, and each sensor runs on a single locking cable. Our LiDAR vs camera FAQ explains why LiDAR is used where precise distances and volumes matter.
Built for the outside of the machine
No cabinets: every computer was installed in the open, so the hardware itself had to provide the protection.
-
01
IP67 and IP69 sealing
A hard-anodised aluminum housing keeps out dust, water and chemicals, and withstands high-pressure wash-downs.
-
02
Fanless, no moving parts
Passive cooling and a solid-state design suit 24/7 operation under repeated shock and vibration.
-
03
M12 locking connectors
Power and communication run through screw-locking M12 connectors that hold through vibration.
-
04
Shock and vibration tested
Endurance tested for shock to EN 60068-2-27 and vibration to EN 60068-2-64, from 5 to 2000 Hz.
-
05
Variable vehicle power
Wide DC input with ignition control, for the different power systems across the vehicle range.
-
06
Wide temperature
Rated for the heat and cold of year-round outdoor work.
IP69 adds protection against close-range, high-pressure, high-temperature water jets; our IP69K FAQ explains the rating, and our rugged computer explainer covers how sealing, cooling and shock testing fit together.
Syslogic RPC Compact A2N
The Jetson TX2 NX members of the Syslogic family used on the project, with dual M12 GbE, isolated CAN FD and IP67 sealing.
NVIDIA Jetson TX2 NX 4GB Rugged Waterproof Computer with M12 Dual GbE
- AI
- 1.3 TOPS
- SOM
- NVIDIA Jetson TX2 NX 4 GB
- Ethernet
- 2× GbE
NVIDIA Jetson TX2 NX 4GB IP67 Waterproof Computer with M12 and 4G/Wi-Fi Option
- AI
- 1.3 TOPS
- SOM
- NVIDIA Jetson TX2 NX 4 GB
- Ethernet
- 2× GbE
What changed on the machines
With sealed Jetson AI computers on board, the customer's code ran on the vehicles in the conditions they work in.
-
Outcome 01
Early warning of danger
LiDAR volume analysis flags people, vehicles, spillage and ground loss before serious damage occurs.
-
Outcome 02
No external cabinets
IP67 and IP69 sealing let the computers mount in the open and be washed down with the vehicle.
-
Outcome 03
Extra sensors, no extra supplies
Internal PoE boards power additional LiDAR sensors over single M12 cables.
-
Outcome 04
Built for 24/7
A fanless, solid-state design passed shock and vibration endurance testing with ease.
Construction AI FAQs
-
How Is AI Used in Construction?
AI in construction uses machine learning on drawings, schedules, camera, LiDAR and machine data to make projects safer and more productive. Common uses include planning and estimating, detecting people near moving equipment, tracking progress against plans, monitoring stockpile volumes, guiding autonomous machinery and predicting when machines need maintenance.
-
What Is LiDAR?
LiDAR (light detection and ranging) is a sensor that measures distance by firing pulses of laser light and timing how long each one takes to bounce back. Repeating this many thousands of times a second builds a 3D point cloud of the surroundings, which machines use to see shapes, distances and movement in real time.
-
LiDAR vs Camera: What Is the Difference?
LiDAR measures distance directly with laser pulses, giving accurate 3D depth in any light. A camera captures color, texture and fine detail at low cost, but needs light and has to work out depth indirectly, through stereo vision or AI. Many autonomous machines use both, combining the camera’s detail with LiDAR’s depth.
-
What Is Autonomous Mining?
Autonomous mining uses driverless haul trucks, drills, loaders and dozers that operate without a person in the cab. The machines combine GNSS positioning, LiDAR, radar and cameras with onboard computers and a site control system, so they can haul, drill and load around the clock while keeping people out of hazardous areas.
-
What Is Sensor Fusion?
Sensor fusion is combining data from several sensors, such as cameras, LiDAR, radar, IMUs and GNSS, into one more accurate and reliable picture than any single sensor can give. Autonomous vehicles, robots and drones use it to understand their surroundings and their own position.
-
What does IP69K mean?
IP69K means an enclosure is dust-tight and protected against high-pressure, high-temperature water jets per DIN 40050-9 — 80°C water at 8—10 MPa from close range. It is the food and beverage industry standard for sanitary wash-down on both sides of the Atlantic — and it does NOT imply submersion protection.
-
What Is an M12 X-Coded Connector?
An M12 X-coded connector is the 8-pin M12 coding for Gigabit and 10 Gigabit Ethernet. Its four twisted pairs are separated by a cross-shaped shield that controls crosstalk at high speed, and it can carry Power over Ethernet with a suitable cable. It is defined in IEC 61076-2-109.
Putting AI on a construction machine?
Tell us about your vehicle, sensors, and power, and we will spec a rugged Jetson platform built for shock, vibration, and temperature.