Connect with us

Interviews

Alex Sappok, CEO of RaySecur – Interview Series

mm

Published

 on

Alex is the CEO of RaySecur, he was previously Founder and CEO of FST, an MIT spin-out company developing advanced sensors acquired by NYSE: CTS. He holds over a dozen patents in the fields of radio frequency sensing and S.M. and Ph.D. degrees in Mechanical Engineering from MIT.

Raysecur has developed breakthrough imaging technology based on millimeter waves. The MailSecur systems provide real-time, 3D imaging to detect even the smallest threats – a vast improvement over traditional static 2D X-ray scanning.

What initially attracted you to AI?

I was initially attracted to AI partly out of curiosity for what it could enable and partly out of necessity as a tool for solving complex technical problems.  At the time I was leading a team developing a new type of sensing technology with broad applications – from cars to chemical plants.  Calibrating these sensors for a particular application was very time consuming and involved collecting large amounts of real-world data and then manually processing the data to arrive at a calibration.  AI approaches enabled a drastic reduction in calibration time and resulted in improved measurement accuracy, while being flexible enough to adapt to the diverse range of applications for our sensors.

Since October 2019, you’ve been the CEO of RaySecur Inc, could you share with us what precisely Raysecur is?

RaySecur is an early-stage, venture-backed, technology company based in Boston.  Our mission is to develop safe, effective, and scalable imaging technologies to “see inside” items.  Think all of the benefits of x-ray imaging without any of the drawbacks.  Using very high frequency radio-waves (terahertz) we are able to develop scanners to see inside items, such as the contents of a box, for example.  Unlike x-rays which use harmful ionizing radiation, terahertz imaging is completely safe.  The fact that it’s safe opens the door to a whole range of possibilities from real-time 3D imaging, to the ability for people to interact with an item as they are seeing the contents inside of it.

Our flagship product, MailSecurTM, is a desktop-size scanner that looks something like and office copier or printer, but allows you to see inside of items.  Anyone can simply hold an item in the field of view and see a live image of the contents inside of it.  Initial applications of the technology include scanning mail for threats.  Today our systems are used by many of the world’s largest Fortune 500 companies and governments, and even to scan mail for heads of state.

Outside of our core enterprise applications, I’m always amazed at the wide range of uses for our systems.   These have ranged from home offices and vacation homes for high-profile executives, to scanning mail-in ballots for threats in the 2020 presidential elections, as well as scanning mail on the campaign trail for presidential candidates.

What are some of the risk factors for mail-borne threats?

The biggest risk is the fact that there is very little security when it comes to physical items coming into a building.  If you think about it, it’s pretty much impossible to walk into a building today without going through multiple layers of security – whether its checking in at the front desk, badge access points, or security guards.  At the same time, every email you send is scanned multiple times for viruses, malware, and other cyber threats.  Yet when the mail arrives, whether it’s a letter or an Amazon box, it typically gets delivered directly to the recipient without any security screening.  That’s a huge risk.

The problem is actually quite widespread – the U.S. Postal Inspection Service (USPIS) reported around 3500 dangerous mail incidents last year or roughly 10 per day –  yet the issue generally gets little attention.  Large companies, government institutions, and high-profile people are often at greatest risk.  Most threats come in small envelopes or packages, small enough to fit in a curbside drop box.  Just last month the New York Times reported on Dr. Anthony Fauci ending up covered in white powder after opening a letter, despite having a dedicated security detail while working on the coronavirus task force.

What type of threats can MailSecur detect?

Our MailSecur scanners are capable of seeing a wide range of threats including weapons and explosives.  Where the technology really excels is in imaging “soft items” like liquids and powders which make up chemical and biological items as well as hoaxes.  In fact “white powder” threats, like the letter received by Dr. Fauci last month are some of the most common threats received in the mail.  This is one area where x-ray systems struggle since x-rays are high energy and often penetrate right through powders and liquids making them nearly invisible, especially in small quantities like those found in the mail.

How does MailSecur compare to traditional X-ray scanners?

There are a few key differences.  One of the biggest differences is that MailSecur uses terahertz imaging which is safe, unlike x-ray scanners which use harmful ionizing radiation.  X-ray systems also require radiation permits and certifications before they can be installed, and they tend to be large like the conveyor systems seen at airports, since they contain heavy lead liners to contain the radiation.

X-ray scanners also require trained operators to interpret the static 2-D images generated by the system.  On the other hand, MailSecur provides a real-time, 3-D view of the contents concealed inside an item.  Because terahertz imaging is safe, it also allows you to interact with the item you are imaging – moving it or rotating it to see it from multiple angles, just like you would interact with an object in everyday life.   Powders and liquids are easy to see because they move around in the package.  This makes the scanner really intuitive and easy to use with very little training.

Another key difference with x-ray scanners, is that the MailSecur scanner is small and portable.  Just like installing a desktop printer, the system can be setup and running in minutes plugged in to any standard wall outlet.  Without the need for permits and radiation safety certifications, MailSecur is also much more scalable than x-ray.  In the past large corporations might only have a few x-ray systems at critical sites, leaving the rest unprotected.  MailSecur is inherently scalable, making it easy to secure all of a company’s facilities no matter where they are.

What are some of the machine learning technologies that are used with MailSecur?

We currently have an active research program supported by the National Science Foundation (NSF) to explore a number of machine learning approaches, given the unique nature of our terahertz imaging data set.  In this case we have a dynamic data set, with live terahertz video, as opposed to static images, of concealed items in full-motion.  As part of this work, we are also looking at combining the terahertz video data with contextual information like barcode tracking and optical character recognition from the mail item to get smarter about the potential risk of a particular item.

Some of the ML approaches we are using include multi-category classification to train the systems to recognize certain features in the terahertz data which might be indicative of a threat to flag the item as suspicious.  Given the fact that our systems need to be capable of real-time classification in the field, we have focused on transfer learning approaches to train the system on an initial data set and further supplement this with data as it becomes available from the client.  Lastly, we also utilize generative neural networks to both enhance and tease out features in the imaging data, which are not obvious to someone looking at the initial image, through a generative process.

Our systems contain internal processing capabilities to run some of these more basic models in real-time, but are also internet of things (IOT) devices leveraging AWS cloud services to develop and train the models.  Aside form ML we also apply a number of advanced machine vision approaches to further enhance and augment threat detection.

In 2020, the United States Department of Homeland Security (DHS) designated MailSecur as a Qualified Anti-Terrorism Technology (QATT). How important is this for Raysecur, and what does it imply for the company moving forward? 

Following 9/11, the SAFETY (Support Anti-Terrorism by Fostering Effective Technologies) Act was created by Congress in 2002 to encourage the creation of anti-terrorism products and services. The Act protects sellers of the technologies — and the companies that buy and use them — with critical liability limitations.

For a new technology like MailSecur, achieving US DHS designation is extremely important, not only for the liability protections it provides to us and our customers, but also for the validation it provides in the market.  Specifically, the designation means that DHS has validated the utility and effectiveness of the technology through both scientific studies and field use.

Customers may also be able to attain lower insurance premiums in recognition of this significantly reduced risk. Companies using security technologies that have not been approved for SAFETY Act protection would not experience that benefit.

Is there anything else that you would like to share about RaySecur or MailSecur?

In addition to all the advantages of combining advanced AI approaches with cutting-edge imaging technology, it’s important to also keep in mind the human element.  At the end of the day these systems, whether hardware or software, are tools that help our customers and their employees do their jobs better, more effectively, and most importantly keep them safe.

When it comes to security applications, knowing what to do and how to respond based on the information these new tools provides is as important, if not more important than the tools themselves.  We call this creating actionable intelligence.  It is for this reason, that we have built a team of former military threat experts, who can connect to any of our scanners anywhere in the world, any time a suspicious item is identified.  These experts are able to quickly assess the situation based on the available information to escalate or de-escalate as appropriate to handle potential threats safely and efficiently.

Thank you for the great interview, readers who wish to learn more should visit Raysecur

Antoine Tardif is a Futurist who is passionate about the future of AI and robotics. He is the CEO of BlockVentures.com, and has invested in over 50 AI & blockchain projects. He is the Co-Founder of Securities.io a news website focusing on digital assets, digital securities and investing. He is a founding partner of unite.AI & a member of the Forbes Technology Council.