Radar startups review 2017
As of April 2017
This article is outdated, please refer to https://medium.com/frontier-tech-review/radar-startups-review-e5e171ed6e20 for an updated one.
Radar (microwave sensor) technologies is a costly thing, one probably needs millions to make just a prototype, and even more to reach the market with a market-ready device. On the other hand, radars could be one of the keys, that open the pandora box of autonomy for vehicles and other robots. Elon Musk claims that optical cameras and radars would be enough to make a car autonomous. So do the others — and big companies like Bosch and Delphi invest hundreds of millions to the technology.
The automotive radar market is on its dawn and the stakes are high — its volume is more than USD 3B in 2016 and expected to reach 9B in 2024 according to market researches. Indeed, it’s a huge market regardless self-driving cars or self-flying drones: growing demand of automotive ADAS systems is a very strong driver itself. Today’s estimated ADAS penetration is about 10%, so one day the market might rise 10 times. The one who makes a promising radar technology is awaited in the smart city market (likely the same size as automotive) and even in the whole 25-billion radar market.
Nowadays the market is headed by Bosch LRR/MRR families (among other things used by Tesla), Delphi ESR 2.5 and a few more products. These vendors sell hundreds of thousands units per year. They got pricy R&Ds, more than 15 years of experience and sales . Should one stay away from the rhino? Of course not, and budgets usually don’t allow companies to rest on their own laurels.
Despite the barriers (costly equipment and labor, limited number of component suppliers, established connections in automotive b2b market) there are some brave ones, who try to change the game. The startup nation (I hope you recognised Israel) is the largest radar market contributor. The probable reason is that radar is one of a core military technologies, and Israel is one of the world leaders there. Thus there are many companies and institutes doing R&Ds in that sphere. Anyway, 4 of 7 companies mentioned below are Israeli-born.
What else can we say according to the data? One should probably check whether there is at least couple of millions in the pocket before getting in radars, and the bigger the better — some startups have dozen times more, if we can call such companies startups.
Why radars? Their key features give radars some advantages compared to lidars and cameras, i.e.:
- solid state form factor means robustness and versatility,
- radars work perfectly in direct sunlight or night, they see through dense fog, rain and snow,
- perfect distance and speed resolution.
The typical automotive radar parameters are:
- Max distance, m: 100–300
- Max speed, mph: 200 and more
- Distance accuracy, m (distance resolution): 0,1–1,0
- Data output: 4D, angle (azimuth, elevation), distance, relative speed of an object
- Angle accuracy (angular resolution): 1,0–5,0
- Field of view, degrees: 45–360
To be clear, common radar perfectly complements optical camera — having camera and a radar (probably cameras and radars) is sufficient, no need in other sensors.
So, who’s on board?
Arbe Robotics
Israel, Tel Aviv, founded 2015*
* — as mentioned in crunchbase and startupnationcentral
Arbe Robotics builds a solution for automotive collision avoidance system as well as a product for drones and swarm robots.
Their product is based on a unique mathematical approach and consists of a sensing module, processing unit and a unique patented processing code. The solution is called ultra-low C-SWAP (Cost, Size, Weight, And Power) Radar.
Arbe Robotics mention that they have the industry’s first ever SAR (Synthetic Aperture Radar) to be implemented in affordable automotive device. As a common radar, their SAR has the ability to detect distance, velocity and direction of an object up to 300 meters far in high resolution.
Artsys360
Israel, Holon, founded 2013, 1,8M raised
ARTsys360 develops a solid-state single-unit 360-degree radar sensor for autonomous cars, smart cities and surveillance. Unlike current scanning or multi-unit 360-degrees systems the approach might allow to make either small and affordable device. Reduction in weight and consumption looks very promising for drones.
Their current lineup has RS360–400 and RS500 units which are used for drone detection. It successfully detects human, flying drones, and vehicles within 360 degrees and up to a few hundred meters.
For autonomous cars, ARTsys360 offers a 77Ghz automotive imaging sensor that fits so-called shark-fin enclosure, placed on vehicle’s roof.
Echodyne
USA, Bellevue WA, founded 2014, 15M raised
Echodyne is building a K-band microwave sensor (radar in 24 Ghz range) for any vehicle or drone to let them sense what’s happening around.
Their approach is to bring metamaterials-based (MESA) radar products to the market, which enables the radar to deliver high-performance electronic scanning in a smaller, lighter, and less-expensive form factor opposite to existing phased array radars, that still remain too costly for commercial use. As the company claims, “MESA operates very similarly to a phased array, but at a tiny fraction of the cost, size, weight, and power, making it ideal for all kinds of high performance commercial applications, including radar vision for drones and cars.”
Though, nothing is said about their processing.
Oculii
USA, Ohio, founded 2013, 3,4M raised
As other sensor makers, Oculii focuses on automotive and intelligent transport spheres. Company makes either radars and sensor fusion units — radar and a camera in a box. Oculii uses well-known K-Band (24 Ghz) range, but their radar’s parameters look slightly better than others. Though, nothing is said about unique technologies used.
Netherlands, Eindhoven, founded 2017, 3,4M raised
Omniradar has build a “One-Chip-Radar” operating in 60 Ghz frequency. It may be applied in automotive and drones spheres. The key advantage is the chip itself — where the complete radar including antennas and Analog-to-Digital conversion are integrated on one piece of silicon.
To be clear, Omniradar is building a chip that has the complete set of analog parts onboard, but leaves the processing up to the customer.
Omniradar’s launching partner, Orlaco, is also well known for its automotive and maritime computer vision systems.
Oryx
Israel, Petah Tikva, founded 2009, 17M raised
Oryx Vision develops solid state sensors for autonomous vehicles. Their technology is called coherent optical radar and looks like a way to combine radar and lidar sensors in one. Like a lidar, it uses a laser to illuminate the road ahead, like a radar it treats the reflected signal as a wave rather than a particle. In other words, Oryx might be using radar’s processing algorithms for lidar’s frequencies, what nobody else does.
The laser mentioned is a long-wave infrared laser, also called a terahertz laser after ~1,000 Ghz frequency that it operates. Infrared spectrum is safe for human eyes and very slightly affected by fog or direct sunlight, unlike other lidar systems and cameras.
Vayyar
Israel, Yehud, founded 2011, 34M raised
Vayyar develops radar imaging sensors, that might be used in automotive, drones and several other spheres. Vayyar has gone far since they already have their proprietary SoC, that is used, in example, in Walabot, a size of a smartphone device, that might look through such solid objects as walls. It is available in industrial and DIY versions.
However, Vayyar does not disclose what their key advantages and differences are.
Don’t hesitate to drop me a line if you have something to tell or found an error above. Ivan, koshurinov@gmail.com.
