In our previous blogpost we discussed the role that Artificial Intelligence (AI) can play in asset management and maintenance. The primary goal of Kaios.ai is to structurally improve asset management processes through automation and predictive power by AI. We believe this leads to less waste and more prevention options, but also to a higher quality of maintenance and more control.
Today, we will explain the steps towards a good AI-solution for Kaios.ai. These steps are Capture data -> Process and enrich data with AI -> Share results. These first 3 steps are automation. The next step is prediction, but before you can predict you need to follow the first steps multiple times. Historical data is necessary, you need to compare the past with the present to predict the future…
First the data needs to be captured. There are many different ways and different equipment that can be used to collect data. It is possible to record data with (multiple) HR cameras and/or 360 cameras, on a car or from an airplane. A fairly new technique to collect data is LiDAR. That is a technology that determines the distance to an object or surface through the use of laser pulses. Regardless of which recording method you choose, always keep in mind that the quality of the recording is important for the end result in our solutions, like we stated last time: garbage in means garbage out.
The variety of the sources to collect data, will also affect the results, especially if you want to look back later to predict. The more data points you have used, the broader your insights are on your asset. Of course, only capture what you need, so start with a session to determine what you really need/want to know! All the choices that are made in the first step affect the results. Our AI-specialists can help you determine your Why, What and When questions so you know you are gathering the right data.
Kaios.ai solutions process lots of data and get more accurate every time they see new (good) data. What for a human eye looks like just a picture has a lot more information in it, our models get it out and present what our customer needs. Also if you have historical data you can start comparing situations. We have a couple of solutions that are ready to use, so outcome is clear. For example, our Kaios.ai Road Marking Solution is an innovative, fully-automated AI-algorithm that efficiently detects and classifies road markings. This solution saves a lot of working hours, resulting in benefits for both the company as well as the employees. But sometimes there is a specific question, a new challenge, in our consultancy we are always ready to pick up a new or smaller AI-project.
We have a platform where you can see the results of your processed data in a satellite map or for example in Opentopo, but you can also extract the results to other geo-programs. No problem.
In the future AI will not only automate, it will also predict where problems are likely to occur before they actually occur. So first you build an AI-model that automates, but keep in mind what you want to know about your asset in the future. If you have historical data, you can start comparing the results. You will see if there are patterns recognizable. For example if you look at road maintenance: weather influence or the influence of traffic movement in different seasons. Or, depending on what data capturing source you use, what role the groundwater level our the underlayer of the ground plays in your maintenance process.
Anyone working in the asset management field will immediately recognize the value of such predictions. For one, these predictions enable prevention of problems, which is often quicker and cheaper than repairing when damage has already occurred. Simple, one-time solutions can then be implemented to save a lot of future costs. These possibilities are how AI enables what we call ‘predictive maintenance’. We believe that prevention in maintenance is one of the core concepts that will change asset management in the years to come.
Does your company already collect data? But are you not using any AI-models? Or you just don’t know where to start? We can imagine you are curious to know and learn more about the possibilities your data has to offer. At Kaios.ai, we have a team of AI-specialists ready to analyze your data and talk about these opportunities. Feel free to contact us!