RoadBotics is an infrastructure technology company that uses artificial intelligence to assess roads. But what does that mean?
The goal of AI is to comprehend better how complex thinking truly works. AI aims to understand human reasoning, goal planning, and language as well as to identify how we learn, which is where machine learning comes into play.
Artificial Intelligence in Everyday Life
Whether you have realized it or not, AI is everywhere. This includes Amazon’s “if you like this, you might also like” shopping recommendations and Netflix’s suggestions for shows and movies based upon your previous history. There are also voice assistants like Alexa and Siri, which are continually learning to get smarter to predict and understand human questions and requests. All of these examples are practical and ubiquitous uses of AI and machine learning. Don’t worry, though. They’re not plotting the end of human existence or trying to take over your life. How we use artificial intelligence isn’t like Westworld or iRobot. This is the real world. And in the real world, we need AI to make life bearable.
RoadBotics and AI, Infrastructure Goals
Currently, most governments manually inspect their roads. Visual inspections have been the industry standard, which has worked out fine, but there are some negative factors to consider. They require a lot of time, are often dangerous, and the conclusions turn out subjective. Our RoadBotics machine-learning technology was developed at the Carnegie Mellon University Robotics Institute in Pittsburgh with the explicit goal of giving roadway managers and their organizations’ complete infrastructure transparency.
“This is not all that different from teaching a child to learn. In fact, the processes are relatively similar.”
Benjamin Schmidt, President, and Co-founder of RoadBotics
It starts with labeled image data. The goal of our machine learning technology is to teach a computer program to learn by showing it examples of road images. But first, we have to tell the machine what it should be seeing in those images and how to rate them on our 1-5 rating scale. To do this we begin by constructing labeled images.
Variety is Key
Just like when teaching or learning a new skill, it’s crucial to have a variety of examples. So we label images from different types of roads, different surface treatments, and from different parts of the world, as large a variety of environments as we can manage. This helps the machine to learn generally rather than to learn in a narrow set of contexts.
RoadBotics trains personnel to draw on images with a digital paintbrush to indicate different distresses such as unsealed cracks, alligator cracks, potholes, cold patches, sealants and so on until we have every object in the image categorized like how we want the computer to categorize new images later. Also, a road expert assigns a grade to each image. Therefore each image has both distresses labeled as well as the final desired rating.
When we have many thousands of these images in a variety of conditions, we feed them into an Artificial Deep Neural Network (ADNN) – one of many types of machine learning programs. Our Data Scientists designed the ADNN to take all of the labeled data and produce an algorithm, so when given a new image the ADNN has never seen before, it will also label the image just like the RoadBotics personnel would.
More to Come
Because of the rapid advancements happening in the machine learning through our close collaboration with CMU, we are continuously improving our model’s ability and accuracy. Also, the more data we collect, the more advanced our algorithms become. To that end, our model is always evolving.
We hope this overview of artificial intelligence is helpful for you. If you are interested in using our product, get started today!
* This blog post is adapted from Ben Schmidt’s article, ‘How Does RoadBotics Use Machine Learning (AI) to Assess Roads?‘
