Daemon AI uses real-time computer vision and on-site edge computing to monitor log flow, optimize step feeder performance and help sawmills identify potential issues before they lead to downtime. Image courtesy of Daemon Automation.
In a sawmill, a jammed log, an uneven gap on a step feeder or a misaligned board can disrupt production before operators even realize there's a problem.
Daemon Automation, the company behind Daemon AI, is helping wood products manufacturers address these challenges with artificial intelligence. Its AI-powered vision technology continuously monitors production lines, identifying problems as they develop and allowing operators to respond before they lead to costly downtime. Unlike traditional sensors, Daemon AI is designed to operate in the demanding conditions found in sawmills, where dust, moisture, vibration and changing light can make conventional monitoring difficult.
Combining intelligent video analysis, on-site edge computing and direct integration with plant control systems, Daemon AI is already operating in more than 30 mills across Canada, the United States, Australia and New Zealand.
But bringing AI into these environments requires more than sophisticated software. It demands substantial computing performance, continuous local processing and reliable hardware - all within a solution that makes commercial sense.
To support its growing deployment requirements, Daemon Automation partnered with Quoted Tech to develop a purpose-built computing platform engineered around its industrial AI application.
The Challenge: Automating Industry's Toughest Environments
Wood products manufacturing presents some of the most demanding conditions for industrial automation.
From log yards and debarkers to conveyors and processing lines, equipment often operates in environments exposed to rain, snow, dust, bark and constant vibration. Outdoor operations introduce additional challenges, with sunlight, shadows and glare changing throughout the day.
The material itself adds another layer of complexity. Unlike manufactured components with consistent dimensions, logs vary considerably in size, shape and orientation.
These conditions make reliable monitoring particularly difficult.
Daemon AI automating an outdoor sawmill line in snow, with four camera views sending log loading and gap control decisions to the PLC. Video courtesy of Daemon Automation.
Why Traditional Sensors Fall Short
For decades, sawmills have relied on photoelectric sensors, proximity sensors and lasers to monitor production lines and detect potential problems.
While effective for many applications, these technologies have limitations. A conventional sensor may detect the presence of a log or board, but it cannot necessarily determine whether the material is correctly positioned, whether a gap is developing or whether a blockage is forming.
Dust, debris, moisture and changing environmental conditions can also interfere with detection.
Traditional machine-vision systems offer more sophisticated analysis, but many are designed for controlled lighting and relatively uniform products - conditions rarely found throughout a working sawmill.
Daemon AI addresses these challenges by using artificial intelligence to interpret what cameras see, rather than relying solely on conventional sensors to detect individual conditions.
A New Approach to Automation in Harsh Environments
Traditional sensors typically monitor individual points along a production line. Daemon AI takes a broader approach, using artificial intelligence to understand what's happening across the production process.
By analyzing live camera feeds, the system identifies the location, position and movement of logs and boards, allowing it to recognize conditions that could disrupt operations.
Each deployment is trained using footage from the mill's own production line, enabling the AI to adapt to the site's specific equipment, materials and operating conditions.
With decision times of 10 milliseconds or less, Daemon AI communicates directly with programmable logic controllers (PLCs), allowing equipment to respond in real time.
Mills use the technology to automate log feeding into debarkers and canters, optimize unscrambler pocket fill and identify misaligned boards before they cause downstream jams.
Keeping the Step Feeder Running Smoothly
One application is the step feeder, where logs are separated and introduced to the production line one at a time. Maintaining consistent spacing is essential. Gaps that are too wide reduce throughput, while logs fed too closely together can create blockages and costly downtime.
Daemon AI continuously monitors the position and spacing of logs, communicating directly with the PLC to adjust feeding operations in real time.
By maintaining a consistent flow of material without requiring constant operator intervention, the system helps mills reduce disruptions and make better use of their production capacity.
Achieving this level of responsiveness requires continuous, low-latency AI processing directly at the mill - making reliable on-site computing infrastructure a critical part of the solution.
Daemon AI automating an outdoor sawmill step feeder, tracking log loader, top step and pocket status and sending decisions to the PLC. Video courtesy of Daemon Automation.
The Computing Challenge: Bringing AI to the Mill Floor
Automating these processes requires Daemon AI to continuously analyze multiple high-definition camera feeds, identify changing conditions and communicate with industrial equipment in real time.
Each additional camera increases the computational workload, while continuous operation places sustained demands on GPU performance, power delivery and cooling.
Processing also needs to happen directly at the mill.
By running AI workloads on-site, operators maintain control of their video data and computing infrastructure without depending on cloud-based analysis or a continuous internet connection—an important advantage for remote sawmills where connectivity can be limited or unreliable.
But deploying powerful GPUs in industrial environments presents practical engineering challenges.
High-end professional GPUs can deliver substantial computing performance, but their cost can make the finished solution less commercially attractive. Conventional server configurations may also introduce trade-offs involving GPU selection, chassis design, cooling and deployment requirements.
Daemon AI needed a computing platform that balanced inference performance, hardware cost, thermal management and operational reliability.
Just as importantly, the solution needed to be repeatable. Every new customer installation couldn't become another custom hardware engineering project.
"We designed Daemon AI to solve the vision problems in wood products that others couldn't: difficult scenes, inconsistent camera quality, awkward angles, and especially outdoor control applications. We built it around the cameras mills already have, because they put those cameras up to catch problems in the first place, yet operators are being asked to watch far too many screens at once. Where we're headed is freeing operators from the chair. We'll need skilled operators well into the future, but by augmenting what they can do and alerting them remotely, they can walk the line and take on operator-based maintenance and other work they can't do from a station. Running this across more than 30 mills takes reliable compute on the plant floor, and Quoted Tech is our first choice."
Trevor Brander, CEO, Daemon Automation
The Solution: Engineering Around the Workload
Rather than starting with a conventional enterprise server configuration, Quoted Tech approached the project by examining the technical and commercial requirements of Daemon AI's application.
The resulting design centred on the NVIDIA GeForce RTX 5090, providing substantial GPU acceleration for the parallel computations involved in real-time AI inference.
For Daemon AI, the RTX 5090 offered a commercially practical alternative to considerably more expensive professional GPUs, helping balance processing performance with the cost of the finished solution.
But selecting the GPU was only part of the challenge.
A high-performance graphics processor operating under sustained workloads generates significant heat and places considerable demands on power delivery and cooling.
Quoted Tech engineered a custom 4U rack-mounted system around the GPU, selecting the chassis, power supply, airflow and cooling configuration to support Daemon AI's processing requirements.
The objective was not simply to install a powerful graphics card in a server. It was to develop an integrated computing platform designed for sustained AI inference while remaining practical to manufacture and deploy.
By engineering the system around the application, Quoted Tech helped Daemon Automation establish a consistent hardware configuration that could support its growing commercial deployments.
From Custom Hardware to a Deployment-Ready Appliance
Computational performance was only part of Daemon AI's requirements.
Its technology also needed to integrate with industrial equipment, including programmable logic controllers and specialized machine-vision systems.
These applications required Windows 11 Pro, making operating-system compatibility an important consideration in the hardware design.
Daemon Automation also wanted to minimize the configuration required when installing systems at customer facilities.
To address this, Quoted Tech worked with Daemon AI's software image and pre-imaged the systems before shipment.
Rather than receiving an unconfigured computer and preparing it entirely on-site, Daemon Automation could receive a system already configured with its intended software environment.
This established a consistent starting point for deployment, reducing the need to repeat software preparation and configuration steps at every installation.
For a company deploying specialized industrial AI technology across multiple customer locations, that consistency is particularly valuable.
It helps transform the computing hardware into a standardized component of Daemon AI's offering rather than a separate integration challenge at each mill.
The goal wasn't simply to deliver a powerful computer. It was to help Daemon Automation turn its AI software into a repeatable, deployable solution.
The Impact: Making Industrial AI More Practical to Deploy
The collaboration provided Daemon AI with a computing platform designed around the technical and commercial requirements of its application.
The benefits extend beyond processing performance.
A Consistent Deployment Platform
With a standardized hardware configuration and pre-imaged software environment, Daemon Automation can build on an established design rather than independently sourcing components, integrating GPUs and preparing systems for every installation.
This provides a more practical foundation for supporting existing deployments and expanding into additional facilities.
A Commercially Practical Approach to AI Computing
By using the RTX 5090 in a purpose-built rack-mounted system, Quoted Tech helped Daemon Automation balance GPU acceleration with hardware costs.
Custom chassis integration, power delivery and thermal management support the objective of operating demanding AI workloads without relying on a more expensive professional GPU configuration.
More Focus on Developing Industrial Intelligence
With Quoted Tech handling hardware configuration, assembly, testing and software imaging, Daemon Automation can direct more of its engineering attention toward developing AI models, improving detection capabilities and adapting its technology to different production environments.
For a growing industrial AI company, that division of responsibilities is important.
Daemon Automation specializes in understanding production processes and developing the intelligence needed to automate them. Quoted Tech provides the computing infrastructure designed to support that technology in the field.
The value of the collaboration isn't simply a faster computer. It's a more practical way to deliver industrial AI repeatedly, reliably and at a commercially viable cost.
The Bigger Picture: From AI Software to Industrial Intelligence
Developing effective AI software is only one part of bringing industrial intelligence to market.
Turning that technology into a commercially deployable solution requires computing infrastructure that can perform consistently in demanding environments while integrating with existing industrial equipment.
Daemon Automation demonstrates what's possible when artificial intelligence is designed around the realities of industrial operations.
With its technology already operating in more than 30 mills across four countries, Daemon Automation is helping wood products manufacturers address production challenges that traditional automation systems have struggled to solve.
And the opportunity extends beyond sawmills. The same approach to analyzing industrial environments, identifying problems and responding in real time has potential applications across other manufacturing and processing industries.
For Quoted Tech, the collaboration illustrates how custom hardware engineering can help technology companies bridge the gap between developing sophisticated AI software and deploying it in real-world environments.
Daemon Automation develops the intelligence that makes its technology valuable. Quoted Tech provides the computing platform designed to support its deployment.
The right hardware doesn't just run the AI. It makes the AI deployable.