Member Info for TheLongestShot

I'm not regular then...

I am sure there are companies with ML experience that could build a CNN that groups posters together based on word usage and times of posting, perhaps they could pick up patterns based on whether the share price is rising or falling.

Sophisticated models could even predict the S**t that they will spout next!

Still building model but implies Guardian units ~22,750 e/o Feb. Using kms traveled Mar-Jun 2019 as proxy implies installs ~28-29k e/o Jun. Not enough data to be sure. Update confirms supply chain ok. 14 distis, NTI & CAT now distributing Guardian, so installs could have upside.

Smoke in your eyes again?

Maybe you should run a business employing 250 people, with sales organisations in multiple countries with a large number of expensive engineers not minimum wage shelf stackers.

Our valuable stock is just that valuable remember 1000 items that cost $1000 each is $1 million whichever way you look at it, but you are ab expert moaner and nothing else

https://www.linkedin.com/jobs/view/1739441708/?TrucksFoT

Head of Automotive Integration
Nauto
Palo Alto, CA, USPosted 20 days ago
SAVEAPPLY
Job description
Senior Risk Consultant, Fleets and Insurance

Nauto? is the only real-time AI-powered Driver Behavior Learning Platform able to predict, actively prevent, and reduce high-risk events in the mobility ecosystem. By analyzing billions of data points from over 400 million AI-analyzed video miles, Nauto’s machine learning algorithms continuously improve and help to impact driver behavior before events happen, not after. Nauto has enabled the largest commercial fleets in the world to avoid more than 25,000 collisions, resulting in nearly $100 million in savings.

Nauto is located in North America, Japan, and Europe. Nauto is backed by SoftBank Vision FundGroup Corp., Greylock Partners, BMW iVentures, General Motors Ventures, Toyota AI Ventures, Allianz Group, Playground Global and DNX Ventures.

Nauto is growing fast, but this is still a ground floor opportunity. Management encourages ownership and accountability and is very transparent with what’s going on. Every role has a lot of visibility and challenges and lots of chances to make an impact.

About Risk & Insurance @ Nauto



Mission-driven individuals passionate about transportation risk and fleet safety and keen to untap insights from vehicle & driver behavior data to inform risk reduction strategies. Be part of the team that develops and evangelizes Nauto’s risk prevention value proposition for our enterprise fleet and insurance industry customers.

In this role you will be creating frameworks to uncover and pinpoint areas where Nauto’s technology will save our customers’ bottom line, working cross functionally with multiple groups internally, including Customer Success, Analytics, Sales and Safety, and externally with Risk, Safety and Actuarial experts within our customer base.



Responsibilities

Develop strategy to support Nauto’s customer success efforts with prospects and clients to precisely state impact to bottom line profitability. Support efforts to articulate Nauto’s value proposition to Risk & Insurance depts. at enterprise accounts to gain their buy-in on Nauto impact to enable upsells and retention Partner with captive Risk departments in Nauto’s enterprise clientele to drive deployment of cost savings initiatives, including claims recoveries
Train Nauto sales and services employees on best practices in fleet risk management and how to better quantify Nauto’s impact to risk reduction & safety initiatives. Serve as subject-matter-expert and go-to consultant related to how Nauto reduces total cost of risk for our customers

PM here's the official press release

https://www.audi-mediacenter.com/en/press-releases/mobility-for-megacities-audi-aime-11484

Yes they showed it at CES, and it got mentioned in reviews, but no real detail to go on. Wasn't that the demo where most of the discussion was about playing games. We probably won't see Level 3 for 10 years or so now, so that isn't relevant

Buffet, even if it was going to be us in the tempest, the ongoing engineering fees or licensing money for work on the headset wouldn't be one we are waiting for, but a great thing to talk about, as long as the NDA doesn't make us wait until it is in service!

Shallwe, it just goes to show how complicated it is to get the next generation of VR/AR. It ask has to sit on your face, will need to be wireless so battery constrained and it would be good if it didn't generate enough heat to fry your skin. But for better interactivity it would be good to include your face and expressions in the experience. Eye tracking plus facial camera is required but that doesn't work with goggles on. JC reminded me that in a recent SEE patent they mentioned Microsoft Kinect. For the uninitiated, it isn't a VR headset, it was a add-on for the XBox games console it had a camera plus IR time of flight sensor released in 2010. It enabled games to use a skeletal model to work out your body pose.

Now VR headsets don't do bodies, just head motion. You can add on two hand controllers to identify where your hands are, but not what the hands are doing. It is like wearing mitts.

So back to the connect, wouldn't it be great to see the real 3d pose, with real facial expressions. Now, picture our guardian unit sitting in front of, or above your TV. It can watch your pose and your gaze. Now imagine a video conference everyone is looking at the screen, not the camera, so no one experiences eye contact. We can fix that on software, but wouldn't it be horrible if the lighting wasnt great so the eye detection was a bit iffy so the eyes go dead mid conversation. Apple does this now in facetime, but that is with a close camera illuminated by the bright screen. So SEE's tech important. Now fancy a game. Fast forwarding a few years, TV is now 2 metres wide. Not a 360 degree experience, but if you look just to the left of the screen, the game moves the viewpoint as if you turned your head. It didn't require a joystick, just a natural glance. Now the camera can see you, it can mimic your pose and expressions to the other players. T-shirt and jeans wont cut it, but in the game you appear in costume - coat of armour, spacesuit etc - perhaps even as an animal but with your pose, facial expression and gaze. Now that is far better than VR. And far more saleable. VR is for Teens and gamers. But telepresence, video conference etc is far more accessible . Remember that the Nintendo wii sold to non gamers. This could be ubiquitous. We don't need to build this, we just licence the IP. In this form, it is far more saleable than VR headsets.

Thanks shallwe, I posted it yesterday ish
;)

Senior Test Automation Engineer
Seeing Machines
Canberra, AUPosted 6 hours ago
SAVEAPPLY
logo
2 college alumni work here
Job description
Underpinned by more than 20 years of scientific research into human behaviour in real-time, real-world scenarios, Seeing Machines has developed the world's most advanced AI driven safety technology. This technology is the bedrock that enables new forms of intelligent machines to see, understand, assist and ultimately protect people who are using them. As our world rapidly evolves, our experienced team continue to stay one step ahead with technology that enhances safety across all transport sectors. We are seeking the brightest minds to help us continue to create and deliver technologically advanced solutions.

About The Opportunity

As a key member of our Test Engineering team, you will be responsible for initial concept design right through to delivering validated system test HIL/PIL platforms for use throughout the Design Validation Life Cycle. You will prepare technical documentation and deliver training modules to our Engineering teams as we verify and validate our state-of-the-art Driver Monitoring technology. You will develop testing strategies for computer vision and machine learning algorithms, and guide the continuous improvement of automotive-grade testing processes. You will collaborate with world-class Scientists, Engineers, and the world's leading car makers – as we work together to shape the future of automotive technology.

We Are Seeking People With Proven Experience To
Design, develop and validate the next generation Automated System Test/HIL/PIL platform to support Seeing Machine product line
Design, develop and validate automated test tools for use throughout the Design Validation Life Cycle for Driver Monitoring Systems
Prepare technical documentation and deliver training modules to our Engineering community
Undertake supplier assessment of COTS hardware and software
About You

You will have:
Tertiary qualifications in Electronic or Systems Engineering, Mechatronics or relevant discipline
Industry experience in design, development and validation of Scalable Automated Test Platforms (System Test, HIL, PIL)
Industry experience with markets test tools and working knowledge of leading suppliers, Vector, dSpace, National Instruments
Programming Experience in the following Python, C++, MATLAB, LabVIEW, TestStand, VeriStand, dSpace Automation Desk considered beneficial
Experience working within the Automotive or similar regulated industry
Experience working with camera-based systems (desirable)
To apply, please include an updated resume and a cover letter outlining your interest in the role, and how you could make a difference at Seeing Machines.

Great way to get a few percent more out of the batteries, who knew our eyes were closed so long

Well I have learned a new word

A little outside of our current field of eye tracking but very interesting

https://www.foundry.com/insights/vr-ar-mr/eye-tracking-in-vr

Smokey,

I am probably late in telling you about this mistake, I haven't checked the rest of your calcs

"So that's approx 4000 in 6 months so 666/month.
Extrapolate that from 31st Jan 2020 to 1st July 2020 and you get 20,551 + ( 4 * 666 ) = 23,215"

Feb 2020, March, April, May, June that is 5 months not 4

JC,
Another great idea. Foveated rendering sounds so strange when you first hear of it - but it is based on how our eyes and brains actually work. We dont see in high resolution everywhere - just where we are actually looking, the further from the centre of focus and the effective resolution drops off, and our colour vision becomes less important, in fact at the very edge if our vision, we only really process movement. But our brains fill in the blanks and remember what should be there and if they feel there is movement on the periphery of our vision then they will turn out eyes, then our head to check.

This is quite an old patent 2016, it like an earlier Microsoft patent to predict based on motion where the eyes will be next ensure that the images are ready before our eyes get there. This may be less important now that we have can have preprocessing of images to ensure that we maintain the frame rate, but both processes of advanced pre rendering of images will improve performance, image quality while reducing processing and preserve battery life. This will mean a more realistic experience with less nausea and mental strain.

It doesn't remove the need for eye tracking but builds upon it.

Have SEE been optimising their eye tracking for close, head mounted devices recently, or does the existing tech scale?

Schlem,
Interesting article nice to get some facts on the current specifications.

This is SEYE's hardware today and tomorrow:

The required hardware is relatively simple. Current Smart Eye systems feature a 1-megapixel camera with 1,280-by-800-pixel resolution. It has RGB and infrared (IR) capabilities so that color images can be captured in daylight. At the same time, the active illumination of IR can render detailed images even in pitch darkness without disturbing the driver.

According to Rundberg, next-generation systems will upgrade to 2-megapixel cameras yielding about 1,600-by-1,300 pixels, thus allowing for a wider view while still preserving the necessary detail. Each Smart Eye implementation is customized for the automaker, to allow for either single or dual cameras that can be mounted in the center stack, roof liner, steering column or other locations.

The system is calibrated based on the camera’s precise position to identify where the driver’s head and face are located. “The system can compute where you are in relation to the car,” said Rundberg. “It knows the lens distortion parameters.” Then a combination of classic computer vision and machine-learning work their magic. The computer vision finds key location points to map your eyes, nostrils, mouth and ears. Then the location and rotation of your head are computed over successive frames.

“We can run this on a mid-range ARM core, like from a cell phone from five years ago,” said Rundberg. “The processors cost less than $10 a pop.” The processing and data can be local unless an automaker wants to use the cloud to track a specific passenger from one vehicle to the other. The system becomes powerful when machine-learning is applied. Patterns related to factors such as head position, the duration of blinks, and the distance of mouth openings are used to infer awareness.

“The major components that go into the drowsiness algorithm are related to blinks, eyelid movements and mouth movements – more than head movement,” Rundberg said. “The blink duration gets longer, but also the slope of the velocity of the eyelid gets slower, as you get more tired.” The AI gets better at measuring alertness with more data. As Smart Eye’s algorithms improve, based on input across its growing customer base, the new code can be deployed to cars via over-the-air updates.

We took the system for a test drive at CES 2020 in Las Vegas. Measurements for eye-blink duration were captured at around 180 milliseconds with no discernible latency. As we spoke, mouth-movement data, from 1 to 30 millimeters, were captured in real-time. The camera’s frame rate of 60 frames per second meant a new image-data snapshot was output every 16 milliseconds.

“Ultimately, DMS would appear to be a no-brainer…” says Freddie Holmes from @automotiveworld in a recent interview with SVP & GM Automotive, Nick DiFiore, where they discussed demand for DMS in L2 and L3 automated vehicles. #seeingmachines #safety #dms https://t.co/M5OXFRXXXH

Glandore,

I believe that the contract was for 40k devices to be manufactured in batches. So they have the capability to make more, but we don't want to tie up capital in having too many in stock, especially if the existing customers already have their Guardian devices at their depots to install when the vehicles are available. It is good to know that there are enough manufactured to get us to the YE target. When the stock runs below a certain level, it would be customary to make the next batch JIT, but perhaps they will place the order earlier if it looks like manufacturing is going to be an issue.

Airbus Advances Autonomy Project As Part Of Future-****pit Concept

See the end for a nugget!

The autonomous taxi, takeoff-and-landing (ATTOL) project is the first visible part of Airbus’ work on a so-called fifth-generation flight deck, which is aimed at improving safety by allowing a crew more time for decision-making. As part of the project, a modified A350-1000 performed eight automated takeoffs in December in Toulouse, using an image-recognition system. Landing trials followed in mid-January.
...
***Although no product development has begun, a certification challenge can be foreseen because some algorithms use machine learning (ML). The European Union Aviation Safety Agency recently released the first edition of its “Artificial Intelligence Roadmap,” which begins to answer OEMs’ questions on how to certify an ML-based system.***

https://aviationweek.com/aerospace/airbus-advances-autonomy-project-part-future-****pit-concept? utm_rid=CPEN1000002212056&utm_campaign=23285&utm_medium=email&elq2=de7c4c9cca2d485ca6f70e6f835c63d9

From @tweetermeyer. Look at the video at 9:22.shows ubers BDMS that they implemented after the Tempe crash. Doesn't look like SEE:

George Hotz's transition from "I'll solve self-driving myself" to "self-driving cars are a scam*" has been a thing of beauty to behold. Genuinely in awe of this lad. https://t.co/iNSgEbxoXd

*actually, a profitable robotaxi is a decade a way