Member Info for Aaroon

Angelo presents a deep dive into Occula NPU as well as the fovio chip. Highly recommend to understand the complexity of DMS

https://www.youtube.com/watch?v=ojoMiPQhXbA

All things remaining constant, and going by an estimate of 18 million EBIDTA loss, i would say we need close to 800k units per quarter, assuming no added contribution from fleet. So from roughly 400k now to 800k. Looks achievable.

i think we can conclude with high degree of confidence honda 360 plus is going to use valeo with see dms.

based on the fact that honda 360 is a valeo system. below is the link and a snippet

valeo has made a major contribution to this technological feat by providing the systems enabling safe driving delegation, i.e. all the sensors and control units essential to the level 3 autonomy features. the vehicle is equipped with five 3d lidars (valeo scala™) and two front cameras, which provide total perception around the car. valeo adds the "brain" capable of merging and processing the data collected. the control unit recreates a detailed 360° representation around the vehicle and, thanks to its algorithms, ensures object detection and safety-related functions.

honda was the first automaker to receive certification for level 3 automation under sae j3016™ classification for its “traffic jam pilot” technology on november 11, 2020.

https://www.valeo.com/fr/valeo-recoit-le-honda-supplier-award-dans-la-categorie-developpement-pour-sa-contribution-au-premier-systeme-au-monde-de-conduite-autonome-de-niveau-3-****logue/

Ah crap, just realized why Martin cannot give cars on road. Its because he doesnt want to give pricing info. SEYE possibly pricing things so low, it would be awful

CB complaints make total sense. I had been burnt before investing in a company that had great growth projections and similar reporting issues as smarteye. There are three ways in which SEYE is deliberately obfuscating data

1. Lumping NRE, AIS and royally revenues into a single bucket. It might still be allright, a company can argue we dont want to give away pricing info

2. Reporting design wins on number of models won, this causes all competitors to seem insignificant

3. Instead of minimum guaranteed revenues report on the total lifetime possibility and then add further that if we win more as is usual with the OEM, our pipeline is 2x.

SEYE will surely grow and report better revenues, but with a management that is not transparent and needing to raise more money, very risky

Thanks NathanR999 for the details

Reading the Magna patent for rear view mirror DMS, i am wondering if there is going to be a legal battle going forward. Does Magna have a moat around how DMS is implemented in the rear view mirror.

Gentex is bringing in structured light for in cabin monitoring.

There is still no Occula NPU like chip in Gentex, atleast i dont see Guardian optical acquired at less than 20 mill USD to have such IP. So there are questions around how good this product really is?

Anyone with some knowledge on the patents, i would love to hear your opinions

This is the domain we should be paying much of our attention going forward

Have been going through the gentex quarterly presentations, for the last few quarters. Based on QA, i think the company in confident of awards soon.

Q1 Call

We're working on a couple of launches right now with some OEMs, where we have

business awards. So those would be probably creating material revenue in the -- of that feature set that you saw at CES. That will be the one driving material revenue the soonest. Following after that, obviously, we still think in the next 3 to 5 years, it's going to be the large-area devices and from an overall dollar standpoint, still has one of the largest opportunities that we've seen as a company.

Q2 Call

Yeah, we're still on target for this calendar year with an OEM on driver monitoring and then into 2025 and 2026, there are some

additional projects that are in launch as well. So we're seeing some good momentum, some good progress and using our geographic real estate

in the vehicle to be able to execute that, so great momentum there.

I think it has the opportunity to be material to the overall growth of the business. The beginning, if it's a baseline feature set, it will be slightly below corporate average margin profile because it is more competitive product, but the revenue dollars are definitely significant and

can and will really, over the next few years, provide some tailwind to our growth story

My take on researching the mirror market is Magna is already seeling high end mirrors to BMW, Volvo and VW.

Gentex has a big share of the market with Toyota being the largest customer.

its going to be a tough competition in years ahead

Have they reached 2 mill units?

Even more surprising to me, is that you can win multiple OEMs in a single day or two. I understand the quarter results around the corner but this kind of RNS makes sure we will never know what volumes are attributed to which OEM.

And again why does SEE have to hide the names of the OEMS and SmartEye can declare so many names

Interesting, these are extension contracts it seems for something smart eye won 2020 on wards, except for one new OEM.

If these are extension contracts already going into production this year, what happened to the original volumes from the existing contracts in 2022,2023 and 2024. Assuming two year delay from 2022, a usual timeline for first contracts.

This is exactly why you need KPIs or some other transparent metric from all DMS company.

Parsing through their reports its impossible to figure out whats NRE whats automotive volumes, what contracts never saw SOP

My understanding is avalability of gen3 was not good news wrt selling gen2. As a customer you don't want a two box solution if you can have gen3 with lower power utilization. Given that SEE had large inventory of gen2, they have decided to make it available in latam, apac with a slight delay to clear off the inventory.

However europe gen3 contracts havent ramped up as they expected, but still had to clear the stock for EBIDTA margins get impacted more to scrap the inventory

Good results considering we will have definite uptake in volumes going forward due to

1. Renault going into production later this year upto 100k per quarter

2. VW ramp up

3. suburu going to production

4. BMW rampup to 500k per quarter

Also lest we forget FORD DMS ASP is close to twice the 10 dollar ASP, as this is a full chip fovio solution.

Going through the software releases made for guardian2 notice two interesting things

On software side, there was a new feature added to release 1.7.0 , Heartbeat statistical information. I wonder if SEE can now detect heart beat information from eye pupil. That would be a big deal i think, as to how accurately maybe get to ask PMG next investor meet.

https://seeingmachines-6111666.hs-sites.com/guardian-generation-2-software-version-1.7.0-safe-launch-plan

On the hardware side,

there is a new accessory to Guardian Gen 2, a light bar for attention warning. Designed to meet the European

Commission’s General Safety Regulations for Driver Drowsiness and Attention Warning (DDAW) systems

https://www.autosense.co.nz/_files/ugd/f3cd37_b4d7c987befc400aa47a09d7e8f7eae5.pdf

The light bar is easily attached to the top of the existing Guardian Generation 2 in-cab sensor, using its existing screws. It redirects the light from existing LEDs in the in-cab sensor and accentuates this so it’s visible to the driver.

The technician will then use a dongle to perform the required software update. Overall, the upgrade takes approximately 30 minutes per device and is available for $159.00+GST per unit.

So they might earn some money for the old gen2 systems already sold

Yeah news all around looks very bearish. but on the flip side i know large funds moving capital around looking for growth stocks in this market.

While listening to H1 FY2024 Results , as part of QA PMG mentioned something very substantial. Somehow I had overlooked it the first time.

Based on what he discussed we should be seeing more Simulator/Training contracts dropping before the Colins product is ready in mid 2025.

Here is the transcript from the discussion at 45 mins

https://www.youtube.com/watch?v=ExA4BGslNao

Just over to aviation, we are working with Colins, no issue at all, we have proof of concept that is outlined. Second phase next FY.

Lots of questions regarding other fields of use, simulation and training.

The Colins agreement we have includes field of use , simulation and training

However there are several programs we are continuing to run with other manufacturers of training and simulators..

This is simulation systems going into Qantas, Emirates and others and Colins has full visibility into this. Supportive of it, they do have the first right of refusal but those programs are continuing. We haven’t reported anything meaningful for a while here, but there are numerous programs underway and RFQs in progress.

I like to be able to give specific dates on when they will fall , clearly that I difficult for us, given that is one of the major disappointments that most investors talk about. … it is progressing and we are quite pleased with that

Basic DMS from what PMG communicated means reduced feature set

Unlike the DMS in the instrument cluster or steering wheel, where the LEDs are exposed, the rear view mirror is hard to detect. you would need to shine a bright light on the mirror.

https://www.youtube.com/watch?v=Ab5jDKz-Ayk&t=143s

VW is already in production, but not finding any reviews anywhere

@chutzpahtaker, but why is it SEE is unable to make OEMS realize this fact?

like BMW OEMs will select SmartEye and then come back to SEE after they have a horrible experience.

its on the management to sell the story and make sure this never happens to begin with.

Valeo/ Magna/ CAT they all understand the importance of having SEE in their ecosystems, but how come SmartEye and Tobii sucessfully managed to sell the SW DMS idea to OEMS. I can only think of two things really

1. The story goes well with the SW defined vehicle

2. SmartEye is aggressive and has lied/ distorted the truth as to the real complexities of DMS

Tim pointed out two years back in the white paper, the types of DMS that are going to be in the market.

group them into standalone and integrated.

He made a forceful case why the Occula NPU is needed. Please read the white paper, if you want to know more.

In crux , standalone DMS will not work without Occula NPU or similar base in hardware. Even integrated DMS which might rely on the huge TOPS of underlying processors will not be able to satisfy the badwidth limitations viz processing say 6 different neural nets. However the integrated will do better than standalone.

FOVIO with Occula can acheive an awesome 3.1 ms latency at just 0.03 TOPS for gaze tracking.

Another reason i think SmartEye will not have a one box solution in aftermarket, it simply cannot, while SEE with occula can easily achieve that.

If Tim is to be believed, there is no path for any of the competition to break into the standalone market, viz mirrors, steering column standalone, till they invest into custom Occula like hardware

https://www.embeddedvisionsystems.it/wp-content/uploads/2023/08/The-DMS-Embedding-Challenge.pdf