am everybody, to Hello, at Ram. I our the Thank you, Noam. am am show you technology. I Today, and I going Co-Founder Arbe. CTO the and
radar dimensions and And perfect is automotive. So know refresh ago azimuthalization can great general the the and why the into years It’s very not and [ph] a is high and features. see the directly by measures conditions. doppler in was it It’s automotive has industry. day XX far all away. of extremely in with over has And weather this affordable. it’s it’s sensor technology reliable We a rate. it very industry It’s It that night. targets very It active sensitive. affected long mature this introduced already four time a range into
Now technology this took and edge the brought in performance. we to it what Arbe, we do at of terms
So that within is limit the the industry. the Arbe is provides automotive the best possible radar that technology of
have channels this the process. that and process, we and MIMO we them in So inside this massive of -- we can we we achieved concept that using achieved radar a use can lot means
some we the for we and receive also of is the we version transmitters use XX -- which radar, XX some which we the size and for aim scaled have -- of is radar of receivers. what back is of front transmitters the what car applications, in for XX which aim XX we version, that main down and our channels. So a use And the
It’s equivalent. number Now in performance. means channels channels really -- radar, of
like So of think the you equivalent about camera can it pixels.
with that, get cameras that a want with it the that pixels to this to compare market is or pixels imagine you radars X,XXX on and XXX the So you camera pixels. XX other exist
in than were able breakthrough market and this we the that’s times the available state-of-the-art today get we of to is So XX because performance better achieve. the
this way in ourselves. the is to very by the design -- that had of to very to achieve hard we system to got goal every radar system to. in was is aspect hard this, get and this and had And by there core of we we the component -- the design So innovate order a radar this
transmitter processing. is we which this chipset, So why for the radar own includes the and digital chip our receiver designed a chip a chip,
In have all unit, we silicon processing call IP our we in a this processing is and high performing ideas cost what innovation puts and digital best core. radar which something embedded chip, dedicated a which the is really into efficient that of
to way the the innovate We on physical the radar also of works had point view. from
performance. is So new we we a is part all had modulation. And to achieve of basis our IP core to of that of this invent this type the how high
critical wide for basic range most [ph] avoidance. obstacle from starting is scenarios, applications of just Dissensor a the driving
truck all Dissensor objects out like at around So the you something the and road, things good the the cars -- the in on Radar fall mapping the that the of we maybe that. environment is obstacles, very have understanding of car and front own and stationary your the
radar is that. the good detecting and mapping to sensitive, avoid it’s this, allowing very So them at car the very
this But type the of radars. not all for case is
resolution the create you map. understand to it’s hard a If this environment have very to low radar, really
with this is channel that and is this have because high we counts. technology only high enabled an this innovation So resolution
to technology. you intersection, there for radar if are lane getting you thing lane this trying range It’s similarly, And or long in get very like understand car enabling into hard a have a in our your are with is resolution is task to and an that order the into. to high see we and coming
But also cases. on we are looking edge
solved of higher elevation in or truck block So stuck normal clear, the map if the But cases the is technology you dimension. is is and radar case a a together. we mixed road radar. can It’s bridge this one most hard of the technology, under road understand is the gets truck very the provide also and accuracy edge because the to because with drive for bridge this that there problematic is if can also for our use
cases, know time see it VIU important failing very other is hard road dictator. we very the VIU today of as users. One use the vulnerable and cars is which we is all the
So It’s talk radar in move detect. harder detect to The is and riders to pedestrians slow about motorcycle very motorbike that them signature they speed. are specifically, we and harder low. and
use So a as really motion we detection principle. the can’t
have We to see like this the this can our something, high users. camera those get is and allow to solves car have sensor resolution road technology that the really another that again,
of the this a technology So is demo how really works.
simpler planning now braking doppler and lane you high-resolution You can see high that real-time, we car. in map and decisions we of and the functions navigate prove radar-only really while and really sensor-only is for changing perception, are in of velocity the car perception, called that the And XXX this -- able with application also entire to map create the and the once can to you trajectory degrees able you ADAS can environment in basic get of be we to could smart are like It’s our autonomous map difficult the in -- to the resolution around emergency for the coverage this perception, this car allow doing road stack Localization then drive. Mapping, that this on is most make one technology the driving, course, with for and conscious we Simultaneous applications -- this and possible is considered but world. this can
now it us technology. using market on we advantage the the in to our are core look that delays covers from strong really do and competitive further very and we And and patent the a to we the order time extend our to portfolio the in future. improve improve -- technology, also created we What protects have
core patents. have we So
our which the is that straightforward, How technology, something is not book. we We of the exactly a new processing. do that radar processing take modulation around you around approach have nine out the it’s the granted core radar is and very not patents the do of
array. innovate design you you for small package package to antenna the the together RF area integrate low have We chips, order down, on had How in get a silicon design in that. to the On the small many design and we -- you because cost area. the a system. in to channels package to How how
way the is not some core do IP also to that it’s trivial, we that hard quite So and in regard. have it
Another fundamental radar is to problem, which industry mutual interference. is very
So with other. same the the and frequency can band each radar is always they operate in interfere
have IP invested lot behave, they look very lot strong a have a solution the and be generated is investigating and time effort we problem. this cope in technology they of these in that of we we very how think field, going to system. factors how how one of also of limiting a and on like the And we So how good the to interferences which with affect interference
point cloud, All am when we on innovation we the ready looking the have like have us Now ahead terms milestones the -- the world. representation I forward of richest in for me. physical of talking technology of and is about today, exciting very
means the we the So image. physical this of that basic the radar to principles generate are using
can into us around are and scene take we all from this this map that there insights really know But more not we is secrets information create this the the -- onto We we we information that and know this inside more can can information. and do.
data probably, the production we basis middle are data which the this So of stage and our year, great this into that today is and super -- the rich from a another it’s is cloud for have we resolution point creating that today have is on. this working in is prototype in this will something a that of it And it go we and year. have solution already a we
really get good are the representation, the very is really a world. trying representation And accurate to what next we of
sensors, stacks on LIDAR, other we you which very sensor has to and it its compare So today if rely lot that driving own problems. a is expensive know autonomous
are introducing, LIDAR. than I trying also something get has we are these times to it’s mentioned and its to a now we we LIDAR before almost with projects with And course, advantages are is similar to XX radar that trying solutions. which these cheaper own get of techniques, But the
alternative can to from we which LIDAR, stack to of driving adoption offer as technology this and if detailed future into systems this we reduce then the a that an can to the we is get this So resolution key can close getting believe in the we market. cost and image, what mass get the for high-resolution is radar and autonomous autonomous this driving technology --
this to So critical a see get happen. very -- this stage to as this we make
We on have we are that the radar. also stack in innovating perception the
will image in radar. And How So that know -- image we doing processing. processing AI we digital revolutionized. believe happen same are the thing people that
the based of radar if you can the get you on approach. rate employ there stack the you of if data a processing techniques these are from use data. on -- most model-based today, more So are much But
we understand were also we are to in best what for the have for techniques will on be leaders we have working radar we understand is techniques to AI different and we internally to adapted, it AI run and and should applications are ourselves developed But many leader be And projects that these trying that -- -- see radar. infrastructure data. AI way on we the
we I radar, we So to the edge. perception, to the radar the radar-only best want best are stack, out we radar with best get say. But approach, say, would okay, that the of the taking the
we to Now system. it does what the overall see want
also working fusion are So we application. on a
we adds sure. up see when of radar you the how stack a have sensor full So radar really have and cameras, the it to And to really camera. much car, that’s want the for the
it’s in, to be lighting know conditions, like, critical. going We that poor
perception and camera, a working you you for you on solve decisions will stack Basically, fusing -- solve are So highly use problem. you data into this the and you we our reliable scenarios. that application showing all radar camera really can the problem make plus in the now if get radar
a of our we all into make -- this to Finally, are to really product. we embed have With chipset. product, it next-generation
So, of we course, do always will and this good for is next-generation. it’s a us in this what
edge radar physical the further we in take will AI resolution of resolution native on the to and are approach going processing. to We improve the terms
cost their run to approach. our the and be the non-scalable So on algorithms people computer necessary computer, so the point there can can increase IP put and order data central to already on algorithm on everything it AI which we the the got can run processor, a radar add of in central in not radar cloud, all a
it So scalability improve by performance and flexibility flexibility the -- future. the the cores creates AI in these the edge, to putting and on system
our enhance capabilities. to software-defined want radar also We
the So, proof. in software this market change way today, update the on sensor very an future operates can It’s and radar make the important is the the we ability that see. fly. we to And
it So needs which market is experience on is better that took and radar, times car, on other but manufacturer in than to the car information. use order imagine this the XX it leverage put still the radar any
Tier can make approach, update. our better it the updating this exploiting are software-defined It proof So ourselves we capabilities. can also better. supporting and And with the and market aligning improving future radar by X radar information model to by software the our
cost, to important have most course, automotive know because more is costs, the channels we for the reducing power and looking better cost and the further same the The and industry cost-sensitive very even thing, reduce of on we ways silicon put are further. integration, in next-generation is reducing
are going going do are is we high-channel ways to transceiver. One we that count it to introduce of the a
the and high-channel this In and transceiver provide have we decision technical will a good to receiver today count a performance -- reduce a was in chips, the on time. the which are already we looking put how them improve system. cost help also So the use and at next-generation, transmitter and and together
very see is ahead about excited road are how the of this we So we future. the and us
you everybody you. listening for So thank back now to Kobi and
