Timbeter Blog

Stay informed, improve your practices, and navigate the timber industry with confidence.

Why using AI is better than measuring logs manually

September 30, 2019
Why using AI is better than measuring logs manually
Even though the world talks about AI, ML, blockchain, optimization and digitalization, the timber industry doesn’t seem to keep the pace – old traditions, enough profitable business and availability of the supply seem to be the reasons why there is an urge for real changes. But, clearly the situation is changing, as there is expected growth in demand for timber and new technologies are born every day, disrupting the traditional industries. In this post, we argue why AI is a good replacement for manual measurements.

Photo-optical timber measurement, which is based on AI and machine-learning technology, enables the detection of the contour area of a log under the bark based on at least 2000 points. Timbeter converts this area to a symmetrical circle and based on this process, the average diameter is calculated. The process is designed to measure the log surface area as accurately as possible by converting the irregular shape of the contour area into a perfectly circular image.

AI is a perfect technology for timber measurement, as the task of how to measure logs is well-defined – it is done in the same way every time and there is no need to use any creative problem-solving solutions. This is excellent for AI because machine learning algorithms rely on finding the set of rules for how to complete its job. As logs do look similar in appearance, recognizing them is something that is very much doable. Measuring logs tends to take an individual a substantial length of time. An algorithm can process the whole image at once, much in the same way a person can immediately guess the rough size of a pile merely by looking at it. But instead of vague volume, the exact data is provided: the log count, the diameter of each log and the volume are instantly available. All the data is available on a digital form, so it can be easily shared and controlled by the other parties.

But, whenever the manual measurement is being compared to the AI-based one, some differences may occur. These differences happen simply because logs are never perfectly round or symmetrical. Here, the question is – what is more important if we consider the maximization of the value of each log?

The problem is that new innovative ways for timber measurement are only controlled and validated via old manual methods and rules that were developed more than a hundred years ago. Old methods are much more limited when compared to the new technologies available as they were based on a limited set of samples. Technology enables measuring each log in a much more objective way – can manual measurement really compete with digital where one log is being detected based on at least 2000 points? Clearly, the results based on digital proof and made with the help of AI provide more transparency and better quality control.

In comparison with the case when an individual is measuring logs manually, a photo-optical measurement tool always measures objectively and in the same way. Since the detection is based on machine learning and artificial intelligence, the results are not dependent on the measurer. It is also a fact that some devices give better detection results (a camera with a resolution of 8 megapixels is our minimum requirement), but the detection process always works in the same way.

As a summary – AI is a perfect way for log measurement, because:

  • Significant time saving – 280 000 logs vs 200 per hour 
  • All the measurements are done objectively and based on real log’s surface area
  • Accurate measurement (2000 detection points per log)
  • All data is in digital format

Still measuring your logs manually? Start using Timbeter! To download the app for devices using Android, click here to go to Google Play Store, for devices using iOS, click here to go to the Apple Store.


See it in action

First 10 measurements are free. No credit card required.