By mounting a camera onto a customized FDM 3D printer, and equipping it with a novel image-processing algorithm, the team has created a bot with independent gap-sealing capabilities. According to the researchers, their automated platform could now provide local authorities with a more rapid means of repairing roads, and protecting people from the hidden dangers of unkempt infrastructure.
Revolutionizing road maintenance
Cracks most commonly occur in asphalt pavements or road networks, and they’re often caused by everyday occurrences such as excessive traffic, footfall and weather variations. While such damage continues to be commonplace in many major cities, it has the potential to cause wider journey delays, and even injure the less-observant.
Rather than replace unsound infrastructure, cash-strapped governments tend to use crack sealant for sidewalk and road repair, but the manual process is laborious and potentially dangerous, exposing workers to oncoming traffic. As a result, research into alternatives has taken-off recently, with various AI-controlled platforms being developed.
A team based at the University of California Davis, for instance, has created two different bionic arm-mounted systems, but they’ve both proven incompatible with certain crack types. Other similar machines have also been prone to shaking due to changes in their environment, making them inaccurate and subject to leakage.
By contrast, considerable progress has been made with automated 3D printing technologies recently, opening up the possibility that a correctly-modified system could take on the role of pre-programmed construction worker. In order to test this theory, the joint research team has therefore built a unique 3D printer that’s capable of depositing with precision, regardless of what’s occurring in its vicinity.
Building a paving repair bot
The scientists built their novel platform around a Prusa I3 3D printer, featuring an upgraded mainboard, and a mounted Arduino-compatible module. Once fitted with a camera, the system essentially works by sending captured images via microcontroller to a path-planning algorithm, which independently identifies where and when sealant will be needed.
Using MATLAB software, the team were also able to configure a separate image-processing algorithm that automatically cancels out background noise. In effect, the program transforms crack images into binary form by segmenting its pixels, leaving only superfluous black squares and identifiable white gaps, expediting the machine’s calculation rate.
Putting their rapid new platform to the test, the researchers then attempted to repair a set of cement paving slab lab specimens. During operation, the system proved capable of precisely filling simple gaps with a cold bitumen compound, but it did take 30 seconds to process each image, significantly longer than the average construction worker.
The custom machine was also guilty of over-filling at times, something that could cause accidents if used within end-use scenarios. As a result, the scientists conceded that in future, their prototype algorithm could be further optimized using deep learning methods, to produce a faster and more accurate platform with on-site field testing capabilities.
Asphalt-loaded 3D printers
Although several different asphalt 3D printers have been developed in recent years, none of them have proved efficient enough to warrant their end-use application.
As long ago as 2015, Advanced Paving Technologies launched a kickstarter to fund its novel Asphalt Paving Machine. The device featured Lidar 3D scanning technology, which enabled it to detect upcoming fractures in the road before filling them, in a faster, cheaper and tidier way than conventional manual methods allow.
More recently, researchers at the University of Leeds have attempted to develop tarmac 3D printing robots, that use UGV-captured data to identify potholes, before finding and repairing them. In future, the compact platform could also be fitted to UAVs, as a means of fixing hard-to-access infrastructure in lofty places.
In a similar vein, researchers from University College London (UCL) have created a customized asphalt 3D printer that’s specifically designed to repair potholes. Using their portable new platform, the team have been able to precisely control the parameters of 3D printed asphalt, and optimize the overall extrusion process.
The researchers’ findings are detailed in their paper titled “A laboratory prototype of automatic pavement crack sealing based on a modified 3D printer.” The paper was co-authored by Jingwei Liu, Xu Yang, Xin Wang and Jian Wei Yam.
Are you looking for a job in the additive manufacturing industry? Visit 3D Printing Jobs for a selection of roles in the industry.
Featured image shows broken asphalt resulting in a pothole. Photo via Self Repairing Cities.