Product Features

Q: Does your program rate the street condition 0-100 and classify

A: Our program utilizes a 0-100 scale, classifying conditions based on the identification and evaluation of factors such as the presence and severity of cracks and potholes, their type, and dimensions.

Q: Can the solution (out of the box) capture road measurements such as identifying lanes, number of lanes, AI to determine parking lanes, the width of lanes, the width of roads, and the width of gutters or sidewalks?

A: Currently, Roadly supports PCI index calculation out of the box, following traffic signs identification as the next feature in the roadmap. Estimation of such parameters as a number, width of lanes, width of roads, width of gutters, and sidewalks is possible for an extra cost. Thus, Roadly can identify them using the same data, so the data collection procedure could be performed only once.

Q: It was mentioned there is machine learning involved – how much interaction does this require from our teams for teaching the system? Or is this more done on the development side of things?

A: There is no need for your teams to be involved in the machine learning process for out-of-the-box features. Usually, there is no interaction required from customers to train a NN. Purely this part is done on the development side of MVP.

Q: Does your program recommend street repair treatment and cost?

A: While our program does not directly recommend street repair treatments or associated costs, it does provide indices for each 10-foot road segment, including the area of the segment. If you have an understanding of the average cost for repairing a specific area with a certain condition, you can calculate approximate costs in a spreadsheet using our data.

Q: Does your program also address striping conditions and replacement costs?

A: We are planning to incorporate these features, but they are not currently included.

Q: Does your program identify curb & gutter and sidewalk deficiencies and costs?

Our program is capable of locating these assets but not of automatically recognizing deficiencies. You will, however, have access to all imagery where these assets are visible, eliminating the need for manual video inspection.

Q: Does it layout any plans and/or quantity estimates?

Our program can count any assets that can be specified in natural language. At present, only "point-type" assets are supported in counting. We do not currently count striping, guardrails, or other longitudinal assets, with the exception of cracks.