This website uses cookies. By using the site you are agreeing to our Privacy Policy.


Click to view a special, video that offers a comprehensive overview of the Infrastructure Photo Analysis Cloud Service.

No.07 supports inspection engineers preservesafety by using AI to streamline their work! Infrastructure PhotoAnalysis Cloud Service


Our science reporter’s sneak report INNOVATIONREPORTSimple, accurate automatic detection ofdefects (cracks) as small as 0.1mm using AI!Infrastructure Photo Analysis Cloud Servicedeveloped from the medical field, supports the work of inspection engineers working nationwide by using AI to streamline their work!

  • Meet Our Reporters
  • Our Technology
  • Commercialization and Future Outlook

Meet Our Reporters



Our Technology

Innovation Reporter Ms. Yuko Watanabe talked to Mr. Junichi Makino, Mr. Kohei Sato and Ms. Riko Onodera of Fujifilm about the Infrastructure Photo Analysis Cloud Service, which supports the safety management work of inspection engineers nationwide by using AI to streamline their work!

Ms. Riko Onodera,Industrial Products DivisionFUJIFILM Corporation Mr. Kohei Sato,Group LeaderSocial Infrastructure System GroupIndustrial Products DivisionFUJIFILM Corporation Mr. Junichi Makino,General ManagerSocial Infrastructure System GroupIndustrial Products DivisionFUJIFILM Corporation

[Infrastructure Photo Analysis Cloud Service]Watanabe:The words “social infrastructure” evoke the idea of something essential to our everyday life. So what kind of infrastructure-related service is “Infrastructure Photo Analysis Cloud Service” exactly?

Makino : The “Infrastructure Photo Analysis Cloud Service”is a support service aimed at streamlining the work of engineers inspecting social infrastructure such as bridges and tunnels.Nearly 50 years have passed since the construction and development of Japan’s high-growth period building boom. To prevent accidents due to the aging of those structures, in 2014 the Japanese government legally mandated periodic inspections every five years, but the number of structures to be inspected was enormous compared to the number of engineers able to inspect them. With the shortage of inspection engineers to preserve safety, how to streamline their work became a social issue.

Watanabe : passionSo the passion to resolve this issue and contribute to safety and security was the start of the “Infrastructure Photo Analysis Cloud Service”! Is this a service you created on your own initiative?

Makino : Yes. In Fujifilm, our employees are imbued with the attitude of contributing to the resolution of social issues in various fields with our own technology , and this led to developing a solution ourselves.

Watanabe : I see! When I hear about safety checks on bridges and tunnels I can easily see it as a social issue as I also these structures on a daily basis. Safety is extremely important, so it’s an issue that needs to be resolved!When I hear about these periodic inspections, someone like me immediately thinks of extremely high or other dangerous places, but an inspection is quite a lengthy process, isn’t it?

Sato : That’s right. Close visual inspections are mandated. Inspection engineers must go to the site to be inspected and check the condition with their own eyes. For example, in the case of a bridge, regulate traffic, a bridge inspection vehicle is used above the bridge, and inspection engineers view the underside of the bridge. If an inspection engineer finds a crack, it’s chalked, meaning a chalk line is made alongside it. This is proof that a qualified person has actually viewed the defect in person. Because cracks down to 0.1 mm are seen and lines drawn by hand, this really is a “close” inspection.

Watanabe :Down to 0.1 mm! To find and inspect such small cracks on a large structure like a bridge must require highly honed skills and great concentration…

Sato : That’s right. For inspections, we use this crack scale index, though an ordinary person could not tell the difference between 0.1 mm and 0.2 mm when looking at a crack. But a highly skilled inspection engineer can identify a 0.1 mm crack as soon as they see it. And when they actually measure it, they’re invariably right. It’s an expert technique. Getting back to what I was saying about inspections, when the chalking is finished, a photo is taken of each crack and the whole defect is sketched along with that photo.Next, on returning to the office, the inspector makes an inspection report to the municipality or local body, and that, with photos attached, creates a complete picture. Then checking this with the sketches, the engineer prepares a report based on CAD industrial data drawings showing exactly where the cracks are.

Watanabe : So there’s a series of inspection work steps up to the submission of the report. It’s not only the onsite inspection work but the subsequent office work that takes up the time!

Sato : That’s right. We felt that we needed to resolve the issue that as much time was needed to create the report as for the onsite inspection, which only the inspection engineer could do.

Makino : Yes, work style reform is another major social issue. From that viewpoint, our hope was to help further streamline inspection engineers’ work time.

Watanabe : So “Infrastructure Photo Analysis Cloud Service” resolves the issue of streamlining work!!

Makino : That’s right. By using AI to do such work as the onsite sketching, the attachment of the photos at the office, and the extraction of the chalking and cracks based on that and the digitization, it was possible to reduce the total work time by about 40%.

Watanabe : Using AI for the sketching and office work cuts the time by almost half! That’s amazing!By the way, we don’t usually associate Fujifilm with social infrastructure such as bridges, but were Fujifilm’s unique strengths able to be used in the area of AI?

Makino : Yes, they were. Actually the “Infrastructure Photo Analysis Cloud Service” utilizes research results from Fujifilm’s medical imaging analysis technology. Fujifilm began as a photographic film company and subsequently researched and produced medical X-ray film.

  • About Fujifilm's Core Technologies

Watanabe : I see!! When I heard about the application of X-ray film and other medical imaging analysis technology, I realized this service was developed from Fujifilm’s technologies and expertise! Was the application of medical imaging analysis prompted by the similarity of blood vessels to the shape of cracks?

Makino : The inspiration came not from the fact that they look alike but from the technical concept.It’s about distinguishing precisely what you want to extract from what you don’t. We thought we could apply the system itself of learning and analyzing based on vast amounts of data.

Watanabe : Both medical care and social infrastructure relate to human life, so the idea of precisely distinguishing is a vital factor. In developing the“Infrastructure Photo Analysis Cloud Service” based on the idea of applying medical imaging analysis technology, did you consult with the engineers actually carrying out the inspections?

Makino : We took the input of the inspection engineers truly to heart when we developed the “Infrastructure Photo Analysis Cloud Service”. Before starting development, Mr. Sato and the other developers actually made individual onsite visits in work clothes like this to about 30 companies, including general contractors, inspection companies and construction consultants, to gain a thorough knowledge of the inspection work flow, the nature of the work, how much time it took and other factors.

Sato : While studying the unfamiliar technical building terminology, we found there were lots of difficult things that we couldn’t have imagined if we hadn’t talked with the inspection engineers actually working on the site.Inspection work on site must be completed quickly because traffic restrictions cause such inconvenience to users. And so the inspection engineers are often very busy working nights at the office preparing their inspection reports.The feeling of wanting to help with our technology became stronger after knowing the onsite situation.

Watanabe : So the “Infrastructure Photo Analysis Cloud Service” was driven by this site-first perspective and the enthusiasm of those creating it!

Onodera : Yes. Since its purpose is to support the work of the onsite inspection engineers and make things easier for them, it was designed as a service that would be easy to introduce, simple to use, and with the functions actually needed by inspection engineers.

Watanabe : Can even non-technical people like me readily use it?

Onodera : Of course! Come on and try it with me now.

Watanabe : Yes, please!

Commercialization and Future Outlook

How to use “Infrastructure Photo Analysis Cloud Service” and its Future Prospects

Onodera : First of all, at the regular inspection site, we do the chalking then take a picture of the cracks.AI automatically extracts it later, so it’s no longer necessary to sketch the shape of the whole cracking area as was done in previous inspection work.

Watanabe : It’s good to reduce the amount of onsite work! It used to be a matter of sketching then tracing, which opened up the possibility of errors, but now the process is simplified so it’s much better! By the way, taking photos has always been an essential part of the inspection work. But can inspectors still take their photos with the same cameras as before?

Onodera : Yes. They can shoot with the same equipment they used in the past. For example, in the case of inspection companies, they can automatically generate a complete drawing from images taken with a tough digital camera that can be used under harsh conditions. In the case of companies that do repair work, they sometimes use equipment such as a single lens reflex camera and tripod, and even with that, analysis from the images taken is possible right down to the finest details like the size of fine cracks.

Watanabe : If they don’t have to get new equipment, then there’s not much of an entry barrier for customers introducing this new service!

Onodera : We particularly emphasized the ease of introduction for our customers when we developed it. The“Infrastructure Photo Analysis Cloud Service” is a Cloud service, so you don’t need to buy expensive specialized software or special equipment. As long as you can connect to the Internet with your own PC, you can use continually evolving AI, and after checking the extractions of the cracks and chalking on the preview screen, you can purchase each image at a unit price of several hundred yen. That’s the flow.

Watanabe : It’s good that the introduction cost is low and that you can purchase the images after checking the cracks and chalking extraction results! Could it actually be used on this notebook PC?

Onodera : Yes! It’s easy to operate. Please have a go.First start up the “Infrastructure Photo Analysis Cloud Service” and log in.If you select a folder containing photos already taken, that folder will be uploaded to Fujifilm's Cloud.After that, AI pastes the photos automatically and displays the extracted images of cracks and the chalking.

Watanabe : It’s really easy! It feels like using a photo album Internet service!

Onodera: No specialized study is needed to use the “Infrastructure Photo Analysis Cloud Service”. It’s one of the strengths of the service that anyone can use it straight off. If you’re satisfied with the analysis results displayed here, you push the order button and you will be charged only then and not before. You can also display any size of cracks in 0.1 mm increments, edit the cracks and chalking, including adding or deleting, and calculate the length and breadth of the cracks. Results can be downloaded in the format the customer chooses. Currently, the results can be saved in all three formats — JPG, DXF and CSV.

Watanabe: Preparing reports, repair estimates and other documents is easy using the various data downloaded!

You can also display any size of cracksin 0.1 mm increments Edit the cracks and chalking,including adding or deleting Calculate the length and breadth of the cracks.

Makino : Ease of introduction and operation are big advantages, but as a service for detecting cracks and chalking, the main strength of the “Infrastructure Photo Analysis Cloud Service” is still its detection accuracy. Applying the technology and expertise built up in the medical field, Fujifilm has absolute confidence in the accuracy of this service. And since AI evolves rapidly by learning, as a result of many people using this service, its detection accuracy will improve further in the future. By continuing to evolve the “Infrastructure Photo Analysis Cloud Service”, we believe we can contribute to resolving customers’ issues.

Watanabe : It’s really gratifying that detection accuracy improves as people use the service! The “Infrastructure Photo Analysis Cloud Service” continues to evolve with AI learning, but what about the future prospects for it?

Onodera : Right now we are offering services for the domestic market, but the social issue of safety is the same right across the world. The need for assessments of aging concrete structures also exists overseas, particularly in Europe and the United States. We would like to contribute to the safety of people living in those countries by offering this service all over the world.

Sato : While we kicked off our service with the urgent social infrastructure issue of bridge and tunnel inspection, roads are another vital inspection target. In addition, residential dwellings like condominiums require large-scale repair works every 13 years. We would like to steadily expand the range of this service, a non-destructive and efficient inspection method, so we can preserve the safety of concrete overall. And with the use of drone and other robotic photography growing, we have set ourselves the technical goal of accurately analyzing the images taken with a range of devices to help protect the safety of engineers carrying out the onsite inspections.

Makino : Our technology is designed very much as a support service for inspection engineers. The final judgment is to be made by people with expert skills. In the sense of giving professionals the information that makes it easy to make a judgment call, our service will have a wide application ranging from offering information on time-related deterioration in the case of buildings, to checking whether buildings were built according to specifications when newly built structures are handed over, to predictions of future danger based on data, and suggesting optimal repair methods and materials. In this way, we intend to evolve the new service in the direction of Fujifilm’s healthcare values of prevention, diagnosis and treatment. Of course, this is no easy task. However, if we collaborate with people in various fields and organize our technologies around the desire to resolve social issues, I believe that we can do it and I think we must do it.

Watanabe : Thank you very much for today!

After the Interview

Yuko Watanabe

I found that Fujifilm technology and expertise developed in the medical field has been applied to the totally different field of social infrastructure where it contributes to people living safer lives.
Also, when I actually used the “Infrastructure Photo Analysis Cloud Service”, I felt that the fact that it was developed and designed to be easy to use and convenient so that large numbers of people would be able to use it was because it was a service created by people who care about people.
Fujifilm is doing everything they can with their technologies to resolve social issues.
I’m excited about seeing what Fujifilm does in various fields in the future!

Please have a look at some limited footage that tells us allabout this Infrastructure Photo Analysis Cloud Service.



Meet our Reporters YUKO WATANABE Age24 Scientific Specialty: Robotics

Hobbies include anime, manga, movies, watching TV, visiting sacred spots and karaoke. Has experience in several sports, including volleyball, tennis, basketball and distance running. A real jack-of-all-trades, she's involved in both sports and the humanities. "I'm so glad that what I've learned in science and engineering will be useful to me as a reporter! I'll be giving it my best!"