Volume measurement of inventory at construction site
The need for a more frequent overview is linked to being able to see profitability, with a more up-to-date picture of production volumes, inventory, and the number of tons of goods sold, in direct relation to operational and maintenance costs. There are many different approaches to ensure better inventory control, including the use of cameras, lidar, drones, crane cameras, weights from machines, and scales.
Volume measurement of inventory at construction site
Whether it's a construction process, moving earth masses, or production of stone, gravel, and crushed rock, there is an increasing need for more frequent measurements in the construction industry. This is directly linked to profitability - with more data sources and regular measurements, the aim is to get a picture of what it costs to produce a ton of a given fraction or a piece of road. There are many different approaches to ensure better mass control, including the use of cameras, lidar, drones, crane cameras, weights from machines, and weights from conveyor belts.
In English, construction machines are called "earth moving machines". It's all about moving masses, so the machine data must be viewed in relation to that. This makes it extra useful!
- Aras Koplanto, Operations Manager, Feiring Bruk avd. Mobil
Fremby has chosen to start by collecting multiple data sources simultaneously to examine the discrepancies between them and to obtain a better basis for analysis. A transaction log with data from various sources is also an important step towards introducing machine learning, as it often requires larger amounts of data to provide value.
This is an activity in our product development, which we want to share a bit about as we develop. This concept has come as input from customers who want to see machine data in relation to produced tons. So we have been working on systems to make smarter mass calculations.
Smarter mass calculation
One of the most time-efficient methods for inventory management today is drone surveying. A drone survey takes place in three phases: first with an overflight, then processing to generate a model, and finally an analysis phase with the actual measurement. Initially, images are collected to generate a point cloud and a 3D map of the area.
Once the updated model of the area is generated, a phase of measurement and analysis work can begin. With defined areas for fixed piles of a specific type of material, you will be able to measure changes over time for each overflight.
The most common approach is to conduct an automatic overflight with a 90-degree camera angle. The pilot can also manually fly around a pile and take pictures from the side at various angles using gimbal tilt, for example at 60° and 45° angles. Before generating a model, you can theoretically add as many images as you want from different angles, and one of the most underrated angles is from the ground. With a regular smartphone, you can take pictures or video from ground level to measure the volume of a pile.
A smarter way to achieve increased precision in the data is therefore to supplement the drone images with images from a smartphone for the same generated 3D model. For some locations, there might be so little activity that you can update the image you have along the way. If you sell X number of tons that you know are taken from a specific storage pile, you can simply update the system with the observation "X tons have been taken from pile with ID Y". This way, you can simplify even further.
We believe it will be useful to have tools that allow more people to have a closer relationship with the mass balance. The more people who can register and use their expertise, experience, and different tools to input data, the better.
Multiple data sources for a better decision-making basis
Multiple data sources are needed, including production figures, number of tons per bucket lift, sales figures, drone surveys, mobile measurements, and so on. Additionally, these must be interconnected to provide added value and be more easily accessible. It's useful to monitor fuel consumption from machines, but it provides even more value to get fuel consumption per ton produced.
By measuring more frequently, you will be able to manage production more efficiently, instead of uncritically adding more of a product that is in short supply. By having more and multiple interconnected data sources, you can make better decisions that take into account various factors such as quality, cost, and time. Aleksander, product manager at Fremby, has done some testing and believes that most companies in Norway can optimize their production by at least 20% today.
See how much you have in stock, and what should be produced, get alerts about minimum inventory levels, and include seasonal or sales variations for a more fluid and optimized operation
- Aleksander Johansen, Product Manager, Fremby