RoadBotics is an infrastructure technology company that uses AI to revolutionize how governments and engineering firms make data-driven decisions about their infrastructure assets.

RoadBotics provides governments, engineering firms, and private industries with objective mapping and rating services. RoadBotics’ data helps our clients make data-driven decisions .

RoadBotics simplifies the data collection process by using a smartphone to capture images of the road. RoadBotics then uses machine learning algorithms (AI) to automate the road rating process.

What more information? Read more: How RoadBotics Uses Machine Learning to Assess Roads

The goal of machine learning is to teach a program to learn by example. Our machine learning algorithms are developed to identify different characteristics of the road by looking at labeled image data.

What is labeled image data?

RoadBotics team members draw on road images with a digital paintbrush in which the different colors represent different distress types – for example, one for unsealed cracks, another for alligator cracks, etc. – until we have every distress and object in the image categorized. In addition, each image is assigned a rating by a road expert. Therefore each image has both distresses labeled as well as the final desired rating.

What happens after?

When we have many thousands of these images in a variety of these conditions, we feed them into an Artificial Deep Neural Network (ADNN) – one of many types of machine learning programs. The ADNN is designed by our Data Scientists to take all the labeled image data and produce an algorithm that when given a new image the ADNN has never seen before, it will also label the image just like the RoadBotics personnel would have had they labeled that same image – most importantly, that the final rating produced by the ADNN is the same as the original rating label. This is called training.

It is an extremely complex process involving advanced processors and millions and millions of parameters. To learn more about our process read this blog post here.

1. To get started, you must submit a map of your road network to our Team. We accept most GIS file formats, although we prefer shapefiles or KML files.
2. Our Team will give you a quote based on the size of your road network (in centerline miles). The quote is provided on our standard agreement.
3. Once you sign the agreement, we start our processes and share an estimated delivery date of the road assessment.

You can fill out our Enterprise Program Interest Form here or call us at +1 412-345-3398, and we will get in touch with you to discuss a potential partnership agreement.

Partners receive online access to the Enterprise Partner Program Platform, where they can create and manage all of their clients’ pavement assessment projects.

RoadBotics works with over 250 municipalities around the world. We also work with a growing network of municipal engineering firms through our Enterprise Partner Program, both domestically and internationally.

We are happy to provide references.

• Comprehensive data collection training by RoadBotics Operations Technicians
• AI pavement assessment
• Access to Roadway, our online map, for an unlimited number of users where you can view an inventory of road images for every 10-foot section and intersection-to-intersection segment scores
• Optional compatible GIS file (shapefile, KML, etc.)
• Optional CSV file of segment scores
• Ongoing customer support with onboarding and training

Centerline miles represent the total length of a given road from its starting point to its end point. The number and size of the lanes on that road are ignored when calculating its centerline mileage.

Our technology can collect and assess the lane the vehicle is driving in plus one lane to the left and one lane to the right.

As long as the view of the road is not obscured by a barrier or wider than 3 lanes, we can assess the road with one pass.

Generally, our definition of a centerline mile fits in well with what a municipality defines as one.

We do not collect data in inclement weather.

Leaves during the fall and snow during the winter can block the view of the road. Rainfall can affect data collection due to glare.

The best data is collected from your road network when the road is dry and weather conditions are ideal for gathering high-quality images. Other factors that impact data quality can include the time of day, sunlight direction, and speed.

While collecting image and GPS data, we also record the accelerometer data. However, this data does not go into our AI algorithms for processing. Currently, we only rely on the images and rate the roads based on visible surface distresses. Our team is actively working on incorporating more data sources to allow clients to make more informed decisions on their assets, paving, and maintenance.

The challenges with identifying rideability using phones are:

• Different phones can have varying sensitivity levels.
• Different vehicles have varying suspensions.
• Calibrating every phone for every vehicle is inconsistent.
• Normally people avoid driving into potholes. The rideability data would not account for potholes and major distress that could damage the inspection vehicle.

As a result, the accelerometer data collected on smartphones can lead to poor rideability estimates.

We do not analyze brick, cobblestone, dirt and gravel roads. However, we collect image data of all the roads within a municipal boundary. These roads will show up as “Other Roads” and will not have any ratings associated with them.

Our Team is continuously training our AI to evaluate other pavement surface types in the future.

Our AI can tell the difference between a shadow and a crack in the road.

While analyzing the pavement, RoadBotics’ AI identifies and excludes vehicles on the road. Unfortunately, parked vehicles obstruct the view of the road beneath them and therefore, only the remainder of the visible road is analyzed and rated.

There are no methods for assessing pavement underneath parked cars.

If you have particularly important streets, it may be worthwhile to close it for parking on the days our Team is deployed for a scan.

Our primary focus is on evaluating pavement conditions for our clients.

Some of our customers use the image library of roads to get an overview of additional roadside assets belonging to the community, such as signs, guardrails, and street and traffic lights.

We are constantly training our AI to see more things on the road and right of way. We will offer asset identification in the near future.

We currently do not calculate road width, but it can be done manually using one of the tools on our platform. When estimating for maintenance and paving, we understand the importance of highly accurate width measurements.

In the future, we plan to offer a rough road width estimation. Rough road widths can be calculated from satellite images. While this is not highly accurate, several governments currently use satellite images to calculate rough road widths and find it sufficient for estimating maintenance and paving costs.

Sensor van-based inspections are generally more expensive and take longer processing times. Moreover, since sensor vans need to be driven to the client’s location, van availability can affect road inspection schedules – and delay paving. Due to the amount of data sensor vans collect, processing times are as high as a year! 

Many small to medium sized municipalities have found that most sensor van services provide too much data – more than what is needed to make informed data-driven decisions. RoadBotics simplifies the data collection process by using just one sensor – a smartphone. As a result, data collection can be completed much faster. 

The data is analyzed using artificial intelligence to generate a 5-tiered rating for every 10-foot section of the road. The data can be accessed through our online interactive GIS platform called RoadWay. In addition, we provide GIS and CSV files for use with your existing GIS or Pavement Management System.