Tech companies have invested billions in developing the autonomous vehicle future - from cars to trucks, forklifts, and more. A few of these vehicles can already be found on the market, and have already been put to use. However, despite the complex algorithms and advanced computer vision technologies, naturally, problems arise in the world of autonomous vehicles. Israeli startup Ottopia aims to be a sort of “phone a friend” backup that takes over remotely and helps computing solve the unknown “speed bump”.
Lost autonomous driving? Let a human take the wheel
From forklifts, shuttles, and cabs, to semi-trailers, trucks, farming equipment, and even ships, many vehicles today are being armed with autonomous capabilities. Although, these systems, which are usually developed under lab conditions, are often bombarded with new challenges once out of the lab, like harsh weather conditions (fog, heavy rain or snow); or bad roads, hidden signs, unidentified objects, construction, and a whole multitude of other unknown hazards on the road. Often, these are hazards that a human driver can overcome in no time, whereas the still “young” computer vision technology may be encountering it for the first time. Ottopia’s system wants to combine advanced autonomous capabilities with the good old ‘ten and two’ human.
While chatting with Geektime, CEO Amit Rosenzweig explained that the company’s system enables a human operator to “take the wheel” remotely - no matter the country or continent - “from a small delivery bot to a 500 ton bulldozer.” Ottopia’s customer base mainly includes automotive and heavy machinery manufacturers, as well as key partnerships with leaders like BMW.
You’re not wrong, the system does sound a bit complex, and unsurprisingly, it is. Rosenzweig listed for us a few technological challenges facing development: Minimal latency dynamic video compression on every chipset - for this he explained that the company developed chip-agnostic video compression tech, with minimal end-to-end latency. Another challenge, he added, was optimizing data transmission, with minimal latency, over wireless networks - mostly cellular. He also noted that Ottopia has developed ML-based algorithms that utilize multiple cellular channels at the same time. Or to put it simply: The company’s tech separates the information, and sends it over various cellular networks; making smart decisions on how to send the information 50 times in one second. Other challenges, he explained, were interfacing with different chip makers that work with different hardware - this led the company to develop a flexible software architecture that has been integrated with the different chip makers at the software level.
Hyundai exec joins the board
Ottopia was founded in 2018 by Amit Rosenzweig, previously at Microsoft, and before that he commanded development at the elite 8200 unit; and Leon Altarac, who is no longer at the company. Currently, the company employs a team of 25 in Tel Aviv, and has raised $13 million to date.
The current $9 million round, led by IN Venture, the capital arm of Sumitomo, also saw participation from Hyundai, Maven, and existing investors MizMaa and NextGear. As part of the investment, Woongjun Jang, who leads Hyundai’s autonomous driving center, will join the Israeli company’s board. According to Ottopia, the new capital will be used to increase headcount by the end of the year.
In addition to the round, Ottopia also announced collaboration with Israeli mobility startup Via. The partnership will see Via’s fleet management skills combined with Ottopia’s remote human driver abilities.