Ripcord is a company built on innovation. How else can you explain a well-funded ($75 million in venture investment) start-up entering the mature and rather staid market for document scanning services? When we met with them at AIIM, Ripcord executives discussed the company’s innovations in areas like automated staple detection and removal, a pneumatically driven paper transport, and storing images at 600 dpi [see DIR 4/12/19].
According to Founder and iCEO Alex Fielding, these developments have helped the company (which was founded in 2017) grow to the point where it is on track to capture over a billion pages this year at its Hayward, CA facility. But, that innovation is not enough to get Ripcord to where Fielding wants it to be. So, he recently upped the ante by appointing Barmak Heshmat to the company’s advisory board.
A former research scientist at MIT, Heshmat is an expert in the area of Terahertz imaging. Terahertz imaging is non-optical technology that uses low-frequency radiation to interrogate solid objects. According to the Ripcord press release, “Today, this technology can read through 9-20 pages of a closed book using Terahertz waves. But for the future, it unlocks the possibility of reading entire boxes of paper without removing the lid.”
Now that’s innovation!
“We see ourselves in a race against time,” explained Fielding. “We are running out of time to capture real knowledge that, the only source of which is basically disintegrating. Our goal is to take all that knowledge that exists on paper and make it free and searchable in the cloud.
“With the current technology we have, including using vision-guided robots to create quality line-scans, we believe we have leapfrogged the current capture market 10-fold. We are very happy with what we are accomplishing. We are on track to digitize over a billion sheets this year at our one existing facility with plans to expand into Japan and elsewhere. But, we really feel we need to leapfrog ourselves and provide even faster access to information on paper. Adding Barmak and his expertise in non-optical imaging is the right approach.”
Terahertz imaging has been around for several years and is currently used in fields like astronomy. Heshmat spent six years at MIT as a research scientist focused on advanced imaging and optics with a specialty in photonics devices. One of his accomplishments was helping invent a new method for batch scanning, for which he is listed as an inventor on a patent.
“Leveraging Terahertz imaging, at MIT three to four years ago, we basically invented a method for scanning batches of layered structures, like a stack of papers, all at once, without having to do mechanical separation,” explained Heshmat. “We use a reflection of radiation to gain information on the ink and the white space. We also receive conical information, which can be translated into depth information. So, with Terahertz imaging, we are able to read through books and boxes and separate them into layers.”
OCR can be applied afterwards to recognize the output. “We use advanced OCR that utilizes optimization techniques like being able to detect letters even though some part may be missing or a character might have a shadow,” said Heshmat. “The accuracy really depends on the contrast we are able to get. If the ink is good and there is a strong contrast with the page, we can get very high accuracy. I would say that currently, we can get 100% accuracy down to nine pages. As you get deeper, the signal gets weaker and it becomes harder to separate the characters. But, we’re currently just utilizing radiation from one side. We are developing technology that will enable us to use it from both sides and expect the results to improve. We are also currently utilizing very low levels of radiation that could be increased.”
It’s probably important to note that the radiation being used is not harmful. “Terahertz imaging uses very low frequency, non-ionizing radiation, so it’s very safe,” said Heshmat.
Heshmat, who also recently launched a stealth start-up called BRELYON will work as a technical advisor for Ripcord. “He will help Ripcord incorporate Terahertz imaging technology and guide us along the way,” said Fielding. “We think through batch scanning and some other non-optical imaging methods, we can eliminate some steps we are currently doing manually or with our robots.”
Fielding said that Terahertz imaging technology will initially be incorporated into Ripcord’s processes in the next six months. “Of course, it will be evolving over the next three to five years,” he added. “We will start on objects and small batches, and then graduate to slightly larger batches before getting to the point where we can digitize a box without taking the lid off. That is the longer-term vision.”
“There are several levels of engagement into the Ripcord pipeline,” noted Heshmat. “We might start with just utilizing layer detection to determine how many objects or pages are in a box. Just being able to identify that a stack has 20 pages can be helpful for Ripcord. Going forward, we might start capturing five or 10 pages at a time and further down the line a small box and then a larger box.”
“In the very near term, we think we will be able to implement Terahertz imaging to give us one more accuracy check,” said Fielding. “One of the challenges every service bureau has is that when you ask a customer how many sheets are in a box, they often don’t know. And by the time it is done being scanned, due to variables like human error, five to seven percent of the pages might be lost. We think we can solve that problem with advanced sensors, optics and other methods to make sure everything gets digitized.
“In the future this technology could also be useful to customers, who for whatever reason, want their documents repackaged the same way they came in. With our robots, when we are removing fasteners and putting the pages through a scanner, that doesn’t mean when the paper comes out it’s going to be that easy to reassemble. Terahertz imaging could open the market to more customers with these types of requirements because the documents could be scanned without having to remove staples or take them out of their folders.”
Fielding noted that Terahertz imaging could also help Ripcord gain more work in the area of fragile document scanning. “We are already working with some archivists and government agencies on these types of projects, which is helped by the fact that our current air-powered scanning transport is very safe,” he said. “But, these organizations have books and papers that are hundreds of years old and very fragile. Utilizing Terahertz imaging could entirely eliminate risk. I like to think about what we could have done with documents like the Dead Sea Scrolls if this technology had been around when they were first discovered. It could open whole new doors when it comes to being able to share knowledge.”
Heshmat noted that Terahertz imaging works very well on bi-tonal text but is not sure how it will work on graphical images. He added that the technology is used in other areas like security and even a geology company utilizes it to examine the make-up of rocks to determine how much oil can be found in an area.
Fielding said that Terahertz imaging will become an integral part of Ripcord’s R&D efforts. “We are now on the fifth generation of our robotics workflow and they have all turned out really well. We’ve recently added a sensor, for example, that has the ability to detect a wet signature vs. a photocopy of a signature. But, we have to continue to build more advanced technology, and we are taking that approach with the addition of Terahertz Imaging.” The technology will be incorporated into the Ripcord robots.
Fielding concluded that he is excited to be part of changes currently taking place in the document imaging market. “When I first started in this space, I approached some of the leaders in the document scanning market and asked them for help with designing things like our paper transport,” he said. “To tell the truth, I was pretty shocked when a CTO told me that the company’s engineers were not really working on that type of technology any longer. That was kind of an eye opener for me. I saw it as a symptom of an industry that was ripe for change. We are now really excited to be in the midst of that.”
For more information: https://www.documentimagingreport.com/?p=6267