Satellites collect images that can be used to identify deformations of infrastructure at millimetre level with InSAR analysis. This project investigates the use of InSAR in the monitoring of deformations of bridges. Can InSAR be of added value for a structural health assessment of bridges which, with aging assets over the largest part of Europe, becomes more and more critical? Nowadays the monitoring of bridges is very labour intensive and often not without danger. InSAR may reduce this.
Satellites collect images of an area at a high frequency over time. By measuring phase differences at pixel level of these images, using InSAR (Interferometric Synthetic Aperture Radar) analysis, it may be possible to identify deformations at millimetre level on a large scale, without instrumentation or measurements on site. The department of MOW wanted to investigate whether InSAR analyses could prove a future alternative, or offer added value, compared to the time-consuming topographic levelling measurements they use now at a low frequency of 1 to 3 or 5 years for monitoring bridges. Can InSAR allow MOW to detect deformations earlier? Can InSAR increase MOW’s efficiency? The safety of its portfolio? With these questions in mind, MOW wanted to obtain insights in which type of high resolution radar images at which cost are needed, which algorithms are best used to analyse the images, and which percentage of bridges of various types can be properly monitored and covered with InSAR, ideally in an automated way.
The Programme for Innovation Procurement of the Government of Flanders supported MOW with this innovation procurement, both in terms of innovative tendering methods and financially, and had the procurement prepared by Verhaert New Products and Services. This process (2020) included:
- A needs assessment via workshops with experts and users, identifying the current issues with the monitoring activities and prioritizing the functionalities of the desired solution;
- A prior open market consultation with interested suppliers to examine to what extent they can provide solutions, and how mature, risky, innovative, feasible (in terms of time, budget, etc.), accurate and scalable these solutions are;
- Confidential interviews with suppliers to improve the understanding of the details involved.
The results of this preparation confirmed that InSAR analyses had already been used to monitor individual bridges (250 bridges in a larger area). For the analyses, commercially available historical high resolution images taken in 2014-2016 were used, and freely available radar images within 2015-2022.
The procedure allowed MOW to award the contract to the best supplier, based on tangible evidence and after an expert jury had evaluated the POC’s (February 2022). It gave both MOW as the participating suppliers the opportunity to learn from the POC experience and adapt the tender contract and their finale offers accordingly. It offered the suppliers the opportunity to demonstrate their ability and clarify the needed developments to automate and encapsulate the analyses for a larger scale.
A lump-sum was given to the suppliers to compensate for their efforts. If not for the POC’s, MOW would have granted the implementation contract to another supplier, whose bid seemed to be better on paper, but whose POC underperformed in reality. So instead of launching two different tendering procedures, one for parallel testing, and one for implementation on a larger scale, the whole experimentation trajectory was handled within one procedure.
At the moment of writing (October 2022) the winning supplier Tre Altamira is executing the implementation contract on full scale. The first results are promising, showing on a large scale the possibility to detect small deformations or changes in the behaviour of the bridges, although a full automated monitoring of bridges on large scale continues to stay challenging, especially for bridges where repairs are carried out.
By the beginning of 2023 we hope to gain a clear understanding of the added value of InSAR for monitoring deformations of bridges. Based on the results, an estimate of the required cost to monitor and cover all bridges in Flanders still needs to be made. If enough added value can be demonstrated, and costs are reasonable, we hope to convince our colleagues to apply this technique and to acquire high resolution radar images for the whole of Flanders, allowing them to set up the monitoring technique for all the bridges, and possibly also other structures.
What Makes Your Project Innovative?
- Using InSAR analyses to monitor deformations of bridges automatically at large scale is not yet used as far as we know, and is definitely new to our department. It forced us to gain expertise in this technique in order to understand the results and be convinced of their accuracy and reliability.
- For the participating suppliers the application of InSAR analyses on bridges on such a large scale was also innovative, as they are confronted with requirements of accurate analysis for small constructions within a very large area. They had to incorporate expertise on bridges and bridge behaviour to adapt their analyses accordingly.
- The organisation of an open market consultation prior to tendering, was new for us. The tendering procedure itself, with the application of parallel proof of concepts during negotiations, was atypical and provided some discussions with our legal procurement experts. But it proved to be the right choice to determine the best supplier.
What is the current status of your innovation?
At the moment of writing, Tre Altamira is carrying out the InSAR analyses of bridges within a region in Flanders on a large scale. It becomes clear that on a larger scale there are still limits to the technique. A balance needs to be found between automation and manually sharpening the results. Overall quite good results are obtained and automated reports with detailed information on the movement of bridges over the past couple of years can be made.
Where maintenance works on the bridges are carried out, some manual interpretation of the data seems to be required. Special care is needed where bridges show a large seasonal amplitude in their movement. A big part of the analysis has been done, but the details of the reporting still need to be determined. At the beginning of 2023 we hope to obtain all results and evaluate the project in each of its objectives, including whether it can provide lessons for application of InSAR to monitoring other infrastructures like quay walls or dikes.
Collaborations & Partnerships
- Department of Mobility and Public Works, Xperta, as initiators of the project they brought in all their expertise on monitoring bridges
- Programme for Innovation Procurement: advice and support on procuring innovative solutions, also financial (50%)
- 4 suppliers participating in the POC’s: e-GEOS, Sensar, SkyGeo Netherlands, TRE ALTAMIRA: they brought in their expertise with InSAR analyses for monitoring
- Verhaert New Products and Services: support with preparing the procurement
Users, Stakeholders & Beneficiaries
- The employees of MOW in charge of evaluating the inspection and (topographically) monitoring of bridges and other infrastructures such as quay walls, dikes, pipelines, etc. The InSAR analyses (will) alleviate their work, allowing them to focus (in the future) on bridges which show a deviation in the expected behaviour.
- All 4 companies participating in the POC’s were given the opportunity to apply their InSAR analyses on a novel case (bridges) and optimize them where needed.
Results, Outcomes & Impacts
InSAR seems to be a promising technique, resulting in a very dense mesh of deformation information on bridges, both in time and in space. Taking into account that even more satellites with even better images will be launched in the future, there seems to be many benefits when applying this completely new approach in the monitoring of the deformations of our assets.
With this technology, we expect to monitor the deformations of our bridges and to alert for required maintenance in a better and more cost efficient way than is currently the case, bringing our attention to problematic bridges in a much earlier state. Moreover, the knowledge on the behaviour of the movement of bridges will increase significantly.
If this project can demonstrate the added value of InSAR (based on an expert evaluation) and future costs for implementation remain reasonable, we hope to use the monitoring technique for all our bridges, and possibly also other structures within the ownership of Flanders.
Challenges and Failures
Even though the known bridges with large deformations were detected by all 4 companies invited to run a POC, there was quite some difference in the results. The biggest challenges proved to be the bridges which show large deformation due to seasonal changes, and only one company was able to detect the deformation behaviour of our largest bridge within the given area. This showed how critical the quality of data analysis is and how challenging it is to monitor both small and very large bridges with InSAR.
It seems to be a technological challenge to detect movements for locations which have had (construction) works executed within the analysed time frame. This leads to a required manual action or interpretation of the InSAR analysis. When looking at very large areas with over 2000 bridges, for which maintenance works are regularly carried out, this might result in failure of the technique or large required extra efforts.
Conditions for Success
- An innovation-friendly management and culture, creating room (dedicated budget and team) for exploring new bottom-up initiatives of employees seem to be key, as was present at MOW.
- The team that initiated the project consists of highly motivated and qualified civil servants with a drive to improve their daily activities and look for innovative solutions to increase efficiency.
- The additional advice and support, also financially, of the Programme for Innovation Procurement, was of much value. It de-risked the innovation procurement trajectory and made it more feasible for MOW.
- Joint trust, cooperation and drive to succeed among all partners: The whole trajectory took more time and effort than we could have predicted. Each innovation procurement makes up an intense public-private cooperation project, with many users, stakeholders, partners, etc. They all have to be kept on board and keep a certain steady pace.
The images of InSAR data contain a lot more information than only bridges. They cover the whole area and include information on the movement of dikes, railways, quay walls, pipelines, motorways, buildings, … As such, if we can demonstrate that they are of added value for monitoring bridges, in a cost efficient way, and decide to buy the images in the future and have them analyzed with InSAR, the information on the movement of all these other infrastructures will be available as well. If this works for bridges, it works for other infrastructures, and we might find other parties to cofinance the follow up.
In this regard InSAR technology can contribute to a new Structural Health Monitoring of public assets, alleviating the public monitoring tasks and leading to more efficiency, accuracy, and thus more safety. Meanwhile this project also offers new business opportunities for the companies involved, as many other applications for InSAR can be conceived. This is just the beginning.
When aiming for an innovative approach, it is important to follow administrative paths that allow for such innovations to have the best possible outcome. Differing from more traditional tendering, the price should pay a less significant role and smaller businesses should be stimulated to participate. As it is often difficult to describe the exact outcome of an innovative project, a tender in different stages is advantageous. By running parallel POC’s with different companies, we could de-risk the outcome.
The exchange of data between different public parties (knowledge on maintenance works dimensions, exact locations of the bridges and their supporting structures) and the analysing company is a critical point for success. Therefore it is important to involve all stakeholders and users, from day one and during the whole trajectory. The more automated all information is gathered, stored and shared, the more automated the analysis can be.
Innovation definitely is challenging. It is nice that these challenges keep you on your toes. Also failures and shortcomings can be a useful result of an innovation trajectory. “If it fails to go wrong, it’s not a real innovation”. “Failing forward”. Sharing your experiences (both positive and negative) with as many parties as possible adds only value to the intense trajectory and the results obtained.
Also within other countries and European initiatives we see that the use of InSAR for different applications is being explored. We try to share our experiences with other (governmental) organizations and participate in other initiatives within this sector. It is clear that the opportunities of this technique are explored by others as well.
With this project we hope to increase both the efficiency and quality of structural health monitoring of public infrastructures (in casu of bridges), the safety of which is important to all people.
- Implementation - making the innovation happen
26 January 2023