Skip to main content

RoadBotics clinches Detroit road assessment deal

RoadBotics has been chosen to use their machine-learning technology to assess the city of Detroit’s entire 4,185km road network. The company will work with PlanetM, a Michigan state networking partnership of mobility organisations, educational institutions, research and development groups and government agencies working together in the automotive sector. RoadBotics will provide Detroit transportation officials with its standard Artificial Intelligence pavement assessment as well as a new AI Maintenanc
January 2, 2019 Read time: 3 mins

RoadBotics has been chosen to use their machine-learning technology to assess the city of Detroit’s entire 4,185km road network.

The company will work with 8439 PlanetM, a Michigan state networking partnership of mobility organisations, educational institutions, research and development groups and government agencies working together in the automotive sector.

RoadBotics will provide Detroit transportation officials with its standard Artificial Intelligence pavement assessment as well as a new AI Maintenance (AIM) tool for unsealed cracks. The company says Detroit will be the first city to pilot the AIM tool, which will assist Detroit in taking a preventative approach to road maintenance.

Oladayo Akinyemi, deputy director at Detroit’s Department of Public Works, said the data will be used to expand the city’s road asset database as part of Detroit’s broader data-driven asset management strategy for the City's Right of Way project.

RoadBotics, based in Pittsburgh, emerged from the Robotics Institute at Carnegie Mellon University, also in Pittsburgh. RoadBotics has used the same technology that moves autonomous vehicles - deep learning-based image processing. It incorporates a proprietary app and standard smartphone, placed on the windshield of any vehicle, to collect roadway image data.

The image data is uploaded to the RoadBotics platform where deep learning is applied to isolate the road from other objects in each image, assess the road condition, and automatically generate a condition rating for the road surface.

The objective rating is based on the presence, type and density of the road surface features and distresses that pavement engineers are trained to identify when visually inspecting roads. Finally, RoadBotics renders the complete assessment on its interactive, online mapping platform called RoadWay.

Mark DeSantis, co-founder and chief executive of RoadBotics, said that more than 90 customers in the US are now using company’s technology. In Australia, RoadBotics is working with engineering firm Fulton Hogan to offer its technology across the south Pacific region.

The Detroit contract come as the company raised nearly US$4 million in financing, led by Boston-based Hyperplane Venture Capital.

RoadBotics also recently announce that the city of Savannah in the US state of Georgia will continue with RoadBotics technology to finish their assessment of the rest of the city's 1,125km road network that started last summer. The city also plans to make RoadBotics' assessment of their road conditions available to the public.

For more information on companies in this article

Related Content

  • US ushers in reforms with new transportation bill
    November 9, 2012
    On behalf of ITS America, Paul Feenstra maps out implications and opportunities for the ITS industry. A critical milestone was reached last month when the US Congress passed, and President Obama signed, legislation reauthorising the nation’s surface transportation programmes, breaking a nearly three-year log-jam which had stymied critical transportation reforms and delayed much-needed infrastructure projects. The law, numbered P.L. 112-141 but known as MAP-21 (Moving Ahead for Progress in the 21st Century),
  • Study calls Inrix off-street parking the ‘clear winner’ in US and Europe
    January 21, 2016
    An independent off-street parking benchmark study carried out by automotive technology research firm SBD has concluded that ParkMe, an Inrix company, beat Parkopedia in data accuracy across the key attribute categories in five cities in the US and Germany. According to the study, overall, ParkMe was 12 per cent more accurate than Parkopedia across a set of core attributes that are essential to automakers for customer satisfaction. Most important, ParkMe was 23 per cent more accurate providing the precise
  • Axis Communications launches new deep learning camera
    October 8, 2020
    To add to its broad portfolio of traffic products, Axis Communications has launched the new AXIS Q1615 Mk III Network Camera.
  • EasyMile selected for autonomous shuttle project in Minnesota
    October 5, 2017
    The Minnesota Department of Transportation (MnDOT) has chosen French driverless technology company EasyMile to lead its autonomous shuttle bus pilot project. According to MnDOT, this will be the first cold climate test in the US. EasyMile will use its EZ10 electric shuttle on MnDOT’s pavement test facility, MnROAD. Testing will include how the shuttle operates in snow and ice conditions, at low temperatures and on roads with salt.