Orca AI: Taking the helm from the captain
Orca AI: Taking the helm from the captain
One was born by the Sea of Galilee, the other grew up on yachts, and both of them never left the sea. But it wasn’t until a chance meeting between Yarden Gross and Dor Raviv that Orca AI, "the Mobileye of ships," was born
Yarden Gross grew up on the water. "My family had a tourist boat and three restaurants on Lido beach on the Sea of Galilee. For as long as I can remember I would sit on the captain's lap when he was driving tourists, and he would show me how to make a ship sail," he says. In fact, you could say that he came into the world thanks to water. "My mother, Lisbeth, is a Dutch woman who visited Israel and fell in love with my father, Tzachi, who took her for a ride in his 'Speed Boat' on the Sea of Galilee. At the age of 23, my mother moved to Tiberias and converted to Judaism. She is a tough woman, I learned to swim from the fact that she just threw me into the water. So I also learned that the secret of the sea is to stay calm - to understand what is happening, to understand the currents, and how the sea works. I spent my youth on the Sea of Galilee rescuing people when the wind swept their inflatable rafts to the heart of the lake," said Gross.
Dor Raviv also grew up with the sea - but in salt water. His father is Dr. Amos Raviv, one of the founders of the marinas in Jaffa and Tel Aviv, which he also managed, and was also a director of the Ports Authority and general secretary of the World Marinas Organization. "My sisters and I were born with the sea," he says. "My parents are skippers, and every Hanukkah, Passover and summer vacation my family went on sea voyages.” In the family album from his childhood, he is shown being at sea from the age of 8 months, and as the years go by he is seen growing up on the boat, playing cards with his sisters on board, scrubbing the yacht with them and huddling with them in the cabin before bed. "I remember myself in boats from a young age. I also sailed in a terrible sea, in huge waves, when the wind is so strong that the boat leans to the side, and you sit on the wall, right on the window that is all water. To this day it feels natural to me to sail my house everywhere."
The life trajectories of Gross and Raviv, now 34, have passed through similar stations. Naturally, they served in the Navy - Gross as a diving course instructor, and Reviv as a patrol boat instructor. The two first met only in the corps commanders course, became friends, but from there their life paths diverged again. Gross completed his service, received a bachelor's degree in economics and business administration at the Interdisciplinary Center in Herzliya and founded the startup Engie, which developed a platform for connecting vehicles to smartphones to warn of malfunctions and compare quotes from garages. "We did nice things, but there were problems with the business model, and the venture was finally closed during the pandemic," he says. "But I came out of there with a lot of scars on my back that taught me how to start a company, recruit people and money, build a business model, a good product and a strategy."
Raviv, on the other hand, signed on to extend his Navy service as a professional soldier. "I was a respected commander and 250 trainees passed under me," he says. "Before leaving the Navy, I was told: 'You are also a guide, you are also a skipper, you are also good at writing literature and you are responsible - go write the IDF a training file for operating autonomous vessels. So I met with professionals from the Air Force, the aerospace industry and Rafael, made adjustments to the fleet - mooring, maneuvering, troubleshooting - and wrote operator instructions. I could have continued because the field was very interesting to me, but I made a terrible mistake and went to study electrical engineering at the Technion."
Why?
"Because my dad said there was a lot of money in it. I had a terrible experience there, and I suffered from exam anxiety. For the first time in my life I came across a wall I could not get through, and it broke me. I went into depression, received psychological help, until I left and went to study industry and management instead and that opened up my chakras.”
His next stop was at a startup named Cortica, which developed a face recognition system. "I learned to program myself, and I wrote the code that automated the whole thing," he says. From there he continued to OrCam led by Amnon Shashua, which developed a product that allows the visually impaired to read texts and recognize faces, "and then I met Yarden by chance on the street," he recalled.
What began as a routine conversation of catch up between old friends, later turned into a business collaboration, which later gave birth to the startup Orca AI, which aspires to be nothing less than a "Mobileye for ships." Gross currently serves as the company's CEO, and Raviv is its VP of technology.
"There was hype around autonomous cars back then, everyone thought it was going to happen tomorrow and wanted to be the ones to develop it,'' Gross says. "So in the same conversation between us, Dor and I began to wonder aloud if it was possible to incorporate something similar in the Navy. That's more or less where it ended then, and only the idea kept gnawing in my head."
You met by chance on the street, talked about autonomous ships, and then you went your separate ways?
"We went on with our lives. I left Engie and went to Indonesia with my brother to clear my head and surf. I surfed there all day and saw the very large sea traffic, and suddenly I thought one could bring to the shipping industry the autonomous capabilities and the connection of vehicles to the cloud. When I returned to Israel, my first phone call was to Dor."
Why?
"I remembered a few years earlier he told me he was working on unmanned vessels. He's the only one I knew who was involved in it."
It turns out that the idea was brewing in both of them in parallel channels. "I had a life-changing event on a post-army trip I did with my father in Gibraltar on a yacht," Raviv recalls. "I really remember the night I was alone on the deck, my dad was sleeping downstairs, and I saw a light at the end of the horizon. I did not know if to wake my dad, because maybe it's something dangerous, and on the other hand my dad is almost 70, I cannot wake him for every nonsense that turns out to be nothing. In the end I decided to wake him up, and it was very lucky I did because we were on a direct collision course with a large ship. I had not yet thought of artificial intelligence then, but this experience was burned into my being because of the fear and uncertainty of what to do.”
How did you feel when Yarden picked up the phone and called you with his idea?
"Eureka! It came exactly in time for my 30-year-old pre-midlife crisis. I worked hard, studied hard, did everything I was told and now I come to work every morning and just want to go home. I felt I was working hard for people who have already made it, and now I want to fulfill a dream of my own, to create something myself."
And what did your father say, although he loves to sail, he wanted his son to become an electrical engineer?
"I came to my parents' boat in Herzliya and told them that I was thinking of setting up something of my own with a partner. My father really honestly tried to get me to drop it, because it's dangerous and 'what do you know.' But in the end he softened up and said ‘If you go for it, we are behind you, and we will also support you financially for three months’.
"I felt everything was directing me to do it. I would not have forgiven myself if I had reached the age of 40 without even trying. I also knew that Yarden was coming with scars from his previous startup, and that reduced my fears. I told myself there was a real opportunity here with a good person in all the things that I am not good at, and that we need exactly all the things that I am good at. "
How do you identify a ship 8 km away through the fog?
The dream that the two old friends want to fulfill is the development of an autonomous system for navigating ships at sea. The system developed by Orca, which was established about four years ago, has already conducted practical experiments in Japan. But both admit that at this point the technology is not mature enough for full autonomy in such complex conditions of giant ships, complicated regulation and oceans whose behavior is unpredictable. In the meantime, the startup they set up has developed a system that monitors countless images, videos and data from ships at sea, such as their depth in the water, direction and strength of wind, position, sailing speed, steering angle, sailing direction and distractions in space (glaciers, other ships, etc.). It cross-references and analyzes this data in real time, to provide warnings for unsafe sailing and to forecast hazards. The existing system boasts a prestigious standard from the American Shipping Authority, and in the not-too-distant future it is supposed to promote the sailing of autonomous ships - all with the help of artificial intelligence.
"The world of shipping is very traditional and not data-driven," Raviv explains. "And this is the revolution we want to bring about in the field - to provide data-based insights to improve sailing safety. If Tesla and Mobileye gather experiences from all drivers to produce a collective mind that learns and understands how to drive - we are building a collective mind of sea captains."
According to the European Maritime Safety Agency (EMSA), about 4,000 maritime accidents occur each year, claiming human lives and causing severe economic and environmental damage. Most of the traffic is from merchant ships - an industry that has proven to be especially vital since the pandemic shut down the aviation industry, and escalated the prices of sea freight. "The pandemic also accelerated the process of connecting ships to the cloud," Raviv explains. "It was not like that a few years ago. A ship once transmitted GPS data and that's it, but today every ship has broadband satellite communication and a connection to the cloud. This is what thrusted Orca forward, because with this data the ships can be managed remotely. A ship is no longer an island floating from place to place. Technology has made ships smarter and improved their operation, efficiency, and safety, advancing them toward autonomy. As the aviation world has advanced greatly in recent years: there is a pilot in the cockpit but 95% of the time he does almost nothing but oversee the systems that do most of the work - that's also what will happen in the world of shipping."
So, Orca technology could have saved the Titanic from sinking?
"The Titanic had no system other than a guy on board who had to spot the iceberg in the dark. We have a thermal camera and artificial intelligence that would decipher the photograph, detect that it was an iceberg, and warn of the danger of a collision long enough in advance to change the direction of the ship and avoid the collision."
Is the technology already ripe for an autonomous ship?
"Not yet. At the moment we are only providing insights to the crew on the ship, but later we will make it possible to eliminate certain roles with the help of automation. The first job that will go will be 'the man on the bridge' (watch keeper) - who watches from the deck to detect dangers and warn about them. It is a very Sisyphean role, to stand for hours upon hours and be on the lookout on a voyage that can last even a month. Our system will be able to replace it in a very good way.
"After the man on the bridge, we will automate additional roles, and we will develop a system that by itself will lower the speed of sailing in severe weather, knows how to prevent collisions, etc., until we reach the ultimate goal, which is to create a complete infrastructure for a commercial autonomous ship that will sail safely, efficiently, and also quickly."
Will captains disappear from the world?
"In the last stage the system will be so advanced that we will no longer need people on the ship, and even if there is a captain in it he will only deal with control, not decision making. The ship will navigate itself, and with the help of satellites the fleet manager will control 50 ships."
Even today there is automatic sailing on ships. What is the difference?
"Auto-pilot brings you to the point where the boat needs to go - but does not prevent any accidents. It's like cruise control: it only maintains speed and direction. We are currently focusing on safety issues, but at the same time we are laying the technological foundations for autonomous sailing."
What are the challenges that are unique to the sea?
"Apart from the different response ranges and distances, there is a great challenge in understanding with the help of a camera that a boat is 8 km away from you and not 2 km or 5 km, for example. Because beyond the computational challenge there are parameters that make it difficult. So we had to do a mapping of what a 100-meter long boat looks like in fog, sun, when it's shaking and more, to produce a good enough distance estimate using a camera. In total, we have registered four unique patents, including the collection of data from a ship in a smart way and the filtering of irrelevant background data that leaves only the events of interest for analysis."
Will autonomous ships precede autonomous vehicles?
"The technological problem is easier, so it is more within reach," Raviv explains. "There are fewer extreme cases when sailing on a ship. At sea, the required range of response is 5-12 minutes to prevent an incident. The distances are 20 km, not 100 meters. It's not like on the road, where pedestrians can suddenly appear out of nowhere, for example, and you have a reaction speed of mere seconds. There are still technological complexities, but it is achievable compared to phase five in an autonomous vehicle, which we still do not know when it will happen, if at all."
Mobileye and Tesla, on the other hand, are huge monsters, and you're a small startup.
"The way to beat big companies is to go after a simple and very focused problem," Gross says. "It is possible to reach the state of a ship sailing with a smaller crew thanks to our technology."
"Pilot within two years, autonomous ship within five"
Last February, Orca conducted a sample cruise of the world’s first autonomous merchant ship around Tokyo, Japan, in collaboration with Japanese shipping giant NYK and the Monohakobi Technology Institute in Tokyo. According to the company's reports, the ship sailed for 40 hours (round trip), over 400 kilometers with almost no human intervention, and successfully performed 107 autonomous maneuvers to avoid colliding with 500 vessels that were on its route
How do you carry out such a project?
"We collected data on the ship for a year and a half to study it," Raviv explains. "We put a captain on shore in front of screens, who saw 360 degrees through the eyes of the system with a remote connection to the engine and rudder to make sure everything was fine. In one significant incident the ship did not know how to maneuver the entrance to the pier, so the captain managed it manually from a distance. But the jump from this demo project to a completely autonomous ship is complex and will require more time."
When do you think this will happen?
"In two years we will start the first pilots of a system that replaces the role of the observer on the bridge, and the ambition is to reach an autonomous ship within five years. There are also matters of regulation here, not just technology."
Is there even a market for autonomous ships today?
"There is not yet a customer who says 'I want an autonomous ship'. And that is understandable. The field is very new, there are concerns, there is no regulation, and there are many problems that need to be solved. So, it will come in gradually, and technology will precede psychology: that is, we will see the capabilities in the field before we see large merchant fleets adopting them. But the end is clear: there will be autonomous ships without people on board."
What about insurance companies? They will probably be happy to get a hold of your safety system.
"We are in the early stages of examining a partnership with the global ship insurance company North P&I. But this is a complex area. It takes time to change the thinking patterns of insurance companies, which are relatively conservative entities, and in the marine field even more so. What will make them understand this best is that our system will be installed on many more ships and you will see how much it has lowered the dangerous events and the claims that follow them across the board. Insurance companies need proof on the ground."
"We're not into hedonistic treats. It's a modest business"
Entering Orca’s offices appears like a navigation mistake. They are located in the building next to the central bus station in Tel Aviv, in front of the busy bus bridge that goes up to the station. It’s not where you’d expect to find a high-tech company. But the inner entrance immediately draws the visitor to another world - Bauhaus-style spaces, with the helmsman ships at the entrance reminding those present of what field they deal with here.
What is an innovative startup like yours doing in a place like the Central Bus Station?
"It's on purpose," Gross explains. "We are not Google yet, just a modest startup, and when you come to the office, the employees should feel it too. Inside they have a beautiful office and the best computer and desk, and the atmosphere is good, comfortable and fun. But we do not sit in the coolest towers in Tel Aviv, because that’s not our DNA. And if a potential employee does not want to come work with us because we are not sitting in Hagag Towers, then it is better that they do not join us."
Gross and Raviv like to talk about the unique DNA required to work in a company that already employs 50 people. "Seamen have a certain tapecast, completely different from pilots, for example," Raviv explains. "They have zero ego, because the crew work is very strong and you must put the common good before your personal good. The tradition at sea is that you do not eat before everyone has food on the plate. Because if you do not work as a team, everything will fall apart. When you are on sea voyages into the night, in difficult weather, in an atmosphere of uncertainty, when you have to act under pressure and solve problems as a team - it all shapes you to work together. There is no ‘me’, there is ‘we’."
We're used to seeing the IDF’s 8200 Unit graduates in startups, not Navy commanders. How is your DNA different?
"8200 Unit people are not expected to be excellent crew members, but to understand their field very deeply. But on a ship, the teams are what’s most important, and there is no such thing as soloism. If someone on the ship is unable to do their job, there is always someone else who knows how to fill their place. This is the advantage we bring to Orca, from our experience in the Israeli Navy. The fact that we know what it's like to deal with real problems under pressure in a small team.
"It's a different world and style than other companies. Yarden and I wanted to build a professional team that would lead the company forward with zero ego, bring in professionals and let them run. We recognize our limitations and that is why we are the youngest in the room. Our entire management team is made up of guys with white hair who are our parents' age. We want to see their experience put into effect."
Orca has so far raised $15.6 million from venture capital funds OCV, Mizmaa and Playfair. Its customers include some of the world's leading shipping companies, including TMS Gas, Maran Tankers, and Latsco. The company is already generating revenue, however since it is still in a growth stage, it is still losing money. Orca is now embarking on a second round of funding against the background of the collapse in the prices of shares in technology companies, especially those that are not yet showing profits.
Are you optimistic about the planned round of funding?
“There have been some strategic investors who led the first round, and there are already some interested in the next round,” Gross says. "The company’s revenue has grown a lot in the last year."
The field of maritime transport has been thriving since the pandemic, but from where do you get the confidence that you’ll raise funds in a market where money no longer grows on trees?
"Things can be difficult if they continue to raise interest rates. But VC funds will always raise money and they need to invest it. So company values have gone down a bit, and that's healthy. If you recruit at a ridiculous value and fail to grow the company to that size - in the next round you will get a slap in the face. The values we saw in the capital market last year were insane, and now investors are stopping for a moment to see where it's going, but in the end they still have a lot of money to invest."
The industry is ridding itself of companies full of hot air. Is there gloating in watching others get cleaned out?
"People like to read about the downfall of all these companies," says Raviv. "We were jealous of these people, but now they have fallen and we are all happy. There is gloating, specifically in Israel. But this does not happen to very sober and very realistic companies, who work hard and build a real product with real technology. And that is where we are. Collecting data from the field, processing technologies for ships, and we show that to customers.
"There were those who sold a dream and made a mockery of the market," Gross adds. "But they are not the vast majority of the industry."
Part of the joy in watching others fail stems from hedonism: eye-popping advertising billboards on the Ayalon highway, vacations in the Maldives, and wild parties.
"We have not changed anything. We have never had grand parties and given extreme treats and we do not do that today. We do not have a ramen tap, only a weekly Happy Hour with beers, and once a year we have a fun day within Israel - not in the Maldives.
"Startups have always existed in an environment of uncertainty. This is their definition in the book," adds Raviv. "I'm sad when startups close, because I know what hard work is being done there, so I really can not gloat. But we were never an inflated company. There is a very real business here that works, and modestly."
