Itongadol.-At the Rio Games, the Israeli-developed machine cheerily scanned athletes and visitors and their belongings — including medals
Lisa Dolev has just returned from Rio where she saw many bronze, silver and gold medals. No, she is not an athlete. Her technology was one of many deployed to help secure athletes’ safety at the recent Games there.
Dolev’s Qylatron, is a large honeycomb-shaped machine that aims to increase the safety of large venues such as stadiums, mega events, concerts and amusement parks. The Qylatron was part of the technologies offered to Rio 2016 by the ISDS, an official sponsor and supplier of the Rio 2016 Olympic games. The machine will also be in place during the Paralympic Games, which will be held there later this month.
“We want our machine to be intelligent and smart and serve people by making security less of a hassle,” 52-year old Dolev, who heads the start-up Qylur Intelligent Systems, said in an interview last week, after her return to Palo Alto, California, from Rio. “Security today tends to treat everyone as guilty until proven innocent. In our view people are innocent until proven guilty.”
The security check should not in any way affect the good times people are having, she said.
The machine has hexagonal shaped cubbyholes built around multiple technologies and including thousands of parts and hundreds of thousands of lines of code. It scans for dangerous items and raises an alert when something is out of the norm.
It works like this: After the machine scans the users’ tickets, the visitors shut their bags into one of its five little cubbyholes and retrieve them on the other side, once again by scanning their ticket. The machine scans the content of the bags and if the light is green then the door opens and users can take their bags; but if the bags contain an unidentified object or an object that is not allowed into the venue, like water, then the door will turn purple and remain shut until the issue is clarified by a guard.
Artificial intelligence systems enable the machines to learn to identify items and remember them, and also learn from the experience of other Qylatron machines, gaining more knowledge as they continue to scan, and improving their decision-making abilities as their experience grows, Dolev said.
One feature clients can choose is to set a personal greeting for visitors written on the top screen of the cubby doors: when your ticket is scanned, if you are a VIP or a season ticket holder, the Qylatron can greet you by name, e.g., “Welcome Tom, enjoy the game”; other users would get the generic welcome, or any other greeting chosen by the venue.
Dolev’s machine was placed at the main entrance of the Olympic village, scanning the bags of athletes and visitors. “People said oh wow, this is so cool,” Dolev said. “We had never seen medals before and we found all kinds of medals in the bag, bronze, gold, silver. That was really cool.”
The security was high but the user experience was not spoiled. “And that is what we wanted to attain,” she said.
The daughter of a former pilot in the US and Israeli air forces, Dolev grew up in air force bases and served as a commander in Israel’s air force R&D unit during her army service. At the same time she achieved her Masters in biomedical engineering from Ben-Gurion University of the Negev. After her army service while getting her PhD in biomedical engineering at Tel Aviv University she worked for a number of years for Israel’s Defense Ministry, spending a lot of time dealing with suicide bombers. She then moved to Palo Alto, California, with her family. It was there, in 2004, as she watched the bombing of Madrid’s trains on TV, that she realized there was need for a machine like the Qylatron.
“In that moment I realized that while all eyes are focused on securing air travel, there is a lack of security for people on the ground,” she said. “Trains, stadiums, concerts and amusement parks all these locations that are becoming targets today, were really vulnerable, and no one was paying any attention.”
A second problem she identified was that that there is a lot of security technology available, but it comes in separate pieces. “There are sensors, chemical detectors, x-ray machines all coming from different companies, not in all- in-one machines,” she said.
In addition, companies are focused on one technology and these technologies tend to “stay static, they don’t evolve with changing needs.”
So, armed with pencil and paper, she set out to create a machine that would include a variety of technologies and grow from experience based on machine learning. Israeli engineers Ziv-Av Engineering Ltd. looked at her technical specifications and designs, said it would be challenging but possible, and then helped develop the machine.
Qylur, which launched its first commercial product at the end of 2015, is based in Palo Alto and has R&D facilities in Israel and Palo Alto, and sales headquarters in the UK.
Investors in Qylur include Wall St. financiers, CEOs and principals of companies, individuals, family funds and a private equity firm, Dolev said. The company has raised $32 million to date without VC funding and has just started a “very large” funding round, she added.
“I have a big vision” for the future, Dolev said. “In the next few years we want to focus on becoming dominant in the category of the intersection between customer experience and security. Ours is a fully integrated entry experience solution, and we want to be dominant in that area.”
In addition, Dolev hopes “to spin out and move these intelligent machines forward. We hope to use these socially networked machines, which talk and learn from each other, for additional uses. More smart machines should be popping up in the future, and we will be in a good place to make the most of our knowledge when that happens.”
In case you were wondering, the name Qylur comes from the star-nosed mole, whose Latin name is Condylura cristata.
“I re-engineered the ‘condylura’ to Qylur,” Dolev said. “This animal fuses thousands of sensors simultaneously in its brain to make decisions and one of our objectives technologically is similar: to fuse multiple sensors into a snake decision and in an ongoing thinking way.”