Building an intelligent solution for visual road inspection

Written by Glenn Brouwer | Nov 12, 2020 2:28:03 PM

Many infrastructure owners struggle to maintain their assets at an optimal maintenance frequency due to their reliance on inefficient manual processes. 

In our new ebook "3 successful business cases in infrastructure management" we dive into a real case study from Unihorn, illustrating how artificial intelligence can streamline workflows and improve efficiency.


Together, BrainCreators and Unihorn have built an intelligent solution for visual road inspection, Inspech. The solution provides immediate insight into which road segments are due for maintenance. If the acceptable standard is exceeded, an inspector will be informed by means of an automatic ticket, which is then forwarded to a road worker.

The challenge

There are numerous  challenges that must be addressed in road inspection;

  • Road inspections are labor intensive and time consuming tasks
  • High costs due to the degree of manual work undertaken by road inspectors
  • Subjective results that are highly dependent on individual judgement
  • Difficult to spot road damage at early stage
  • Automating visual road inspections is very difficult due to the amount of damage, variety in pavement and likelihood that damage can occur everywhere

The Approach

BrainMatter powers Unihorn’s solution, ‘INSPECH’, which aims to detect damage to roads using camera images taken by a specifically designed car equipped with a camera set-up for visual road inspection. INSPECH also uses proprietary algorithms that automatically classify road surface segments in accordance with the guidelines of the CROW or Rijkswaterstaat.

This allows the solution to provide immediate insight into which road segments are due for maintenance. If the acceptable standard is exceeded, an Inspector will be informed immediately via an automatic ticket, which is then forwarded to the relevant road authority. This ultimately enables the inspection of roads to occur faster, more cheaply, safely and efficiently. Additionally, Inspectors will no longer be called out and can make decisions from the office.

The results

Road Inspection

BrainMatter views 550 kilometers of captured road video taken by a specially modified car equipped with a camera for visual road inspection.

Damage recognition

BrainMatter learns to recognise 12 segments of road surface conditions based on the guidelines of CROW or Rijkswaterstaat in days versus weeks.

Initiate immediate action

BrainMatter users can define rules to initiate follow-up actions when a damage has been recognized, e.g. a notification or signal to an asset management system.

 

“Thanks to the collaboration with BrainCreators, we have acquired an innovative technology with the BrainMatter platform that enables us to launch our visual road inspection solution, “INSPECH”, together with our partners.” 

Jeroen Bleeker, Manager at Unihorn

 

Learn more about our use cases in our free ebook.