Member Info for johnchucka

Semicast Research@semicast_res

Semicast Research Retweeted Thomas Bamonte
Great link from @TomBamonte. It really isn’t clear how clocking up ever more miles in perfect conditions brings MaaS any closer to deployment at scale. Weather and nighttime are not edge cases but real life. Maybe @AlexRoy144 is Nostradamus! @seeingmachines @SmartEyeAB

Thomas Bamonte@TomBamonte

“Let’s be clear: mainstream driverless cars or autonomous cars — call them what you will — are still a generation or more away, if they are ever allowed to happen at all.”

Heavy Day but Drop over done imo, Guardian Data collection is critical to Fovio development and the customer base is growing (300) along with the unique compounding Data collection in a market place where we have not real competition except Halfords. The cutting edge environment and Data Gold rush that SEE find themselves in along with part shortage seems all par for the rocky running out of pick axes course imo. Obviously someone didn't foresee the acute issues raised in todays RNS but this is a land grab race and predicting the future price of clever high demand kit is going to be a little approximate. Some of us here met KK and we came away with solid opinions on him and all the management so I think they will be as disappointed as we are by this set back.
The other issue raised here is share rewards which I got the impression that the consensus was they were given so key people were rewarded for two reasons -one should we be T/O to ensure they were properly rewarded and 2, for meeting goals, achievement etc. The issues raised in todays RNS are not related to the progress being made imo. The issues appear to be industry wide not specific to SEE. Guardian customers are clearly growing so progress is being made and the critical Data is being collected to ensure Fovio leads the Data race . SEE also are in a race to recruit talent so have to reward appropriately to keep and attract the brightest and the best
The Contrasting opinions of Safestocks and Semicast along with Goldenboy in terms of T/O make for interesting debate as does the volume of trades today. While it's clear Fleet optimism was priced in, being the only owners of Realtime edge case Data is clearly not .

The biggest take away I have from this is Auto/Fovio needs Fleet so treating them as seperate virticals is as misleading as having 0p value atributed to the unique Edge case Data which quietly gives us the edge on SE, which KK expects will lands us Toyata and FCA etal in the coming months.

I assume off road included here but even so it's a big jump.

Less than a week ago KK told us 'of having 10,000 vehicles on the road', today 16,775 collecting data.

SEE FB post

Did you know? Extensive data collection from 16,775 individual vehicles across the globe shows us that our Seeing Machines Guardian driver monitoring system has detected over 3.6 million fatigue and distraction related events - meaning times when tired or distracted drivers had to be reminded to focus on driving. This is life-saving technology.

Semicast Research@semicast_res

Guardian headunit from @seeingmachines is an automotive grade PC. Biggest component cost will be flash memory ICs (or SSD) to store video footage. Cost of flash memory has SOARED in last 12 mos. & issue exacerbated by fall in AUD to USD over same time.

Semicast Research@semicast_res

RNS from @seeingmachines shows adherence to full disclosure. Review into Guardian will almost certainly conclude that if bill of materials costs is higher, then price to the customer goes up. Guardian has NO direct competition at such an advanced level of functionality.

Phil Koopman: Self-Driving Cars Safety Tips: Methods for Testing a Self-Driving car

https://www.youtube.com/watch?v=YuFq1E9uzkw

The Doubling of Fleet in 12 months and the Semicast estimate that 'the second billion will take something more in the order of about 18 weeks is going to make SEE Edge case Data very valuable.

Interested to know why with this EC Data advantage why are companies choosing SE? Does this not put them at potential reputational risk Uber style??

Semicast Research@semicast_res

It took @seeingmachines about 18 years to develop & build the infrastructure to capture 1 billion kms of real world, on road data using Guardian. Semicast estimates the second billion will take something more in the order of about 18 weeks. #gotomarket

hTTps://twitter.com/johnkrafcik/status/1035163309463494657

On the topic of Edge Cases I imagine with the anticipated growth in the Airline industry the safety gains are in the Edge cases, in that 2% on the Distribution curve so if SEE are monitoring pilots in cockpits and simulators we'll be accumulating hours of valuable Data, which again like Auto has a high market value.

Is SEE the only company with this Edge cases Data ? for Level 3/4?

Semicast Research@semicast_res

Even if the answer is 50 billion miles or 100 billion km, Guardian from @seeingmachines will eventually gather all practical edge cases for fatigue & distraction through Human Factors research & real world, on road, detection. What then is the value of that dataset

Mr Semicast

Where does that valuation come from in your tweet?

'So if valuations are at about US$1 billion per million miles of autonomous driving data, what is the valuation for 812 million miles of driver distraction and fatigue data? It clearly isn’t zero'.

Semicast Research@semicast_res

So if valuations are at about US$1 billion per million miles of autonomous driving data, what is the valuation for 812 million miles of driver distraction and fatigue data? It clearly isn’t zero. @seeingmachines

Semicast Research@semicast_res

Real world = not in a simulator
On road = not in a lab
As early L2 systems show it takes about a billion miles of operation to truly understand how humans relate to automated driving features. Maybe 3-4 billion more to understand complex edge cases.

Semicast Research@semicast_res

A Driver Monitoring Engine is the "core signals" data gathering engine common to all verticals. Will be modified slightly for each of the multi-sector segments that @seeingmachines is focused on. Each will have different parameters. FPGA is the PERFECT fit for this strategy.

Semicast Research@semicast_res

Hi @kirstenkorosec @Tweetermeyer @AlexRoy144 This is why L3 WILL happen. L3 is all about Human Factors research. @seeingmachines & @DrMikeLenne have gathered 1.3 BILLION kms of real world driving data for the auto grade DMS. @waymo went to L4 too soon. http://www.seeingmachines.com/wp-content/uploads/2018/08/Seeing-Machines-Guardian-Data-Final-29-08-18.pdf …

Real Time is the most important element here . Data from Real Lives, people in everyday situations.

It's valuable.

Data is Gold rush inside an AV Gold rush. A big piece of real time Data at a time when a lot of companies are driving round suburbia, cities and old military bases and doing computer simulation to catch up and get the edge on the random behaviour of Mr Complexity.

Mike Lenné@DrMikeLenne

First insights from our @actgovernment CANdrive automated vehicle trial presented @ITS_AUSTRALIA #ITSSummit. A question. Automated driving is on. Hands are on Wheel. Eyes are off Road. Is this driver attentive? Steering-based sensors may say YES. Our DMS says NO #drivermonitoring

Semicast Research@semicast_res

Auto grade camera-based DMS design-wins look set to be contested by: Aisin Seiki, Clarion, Denso, FotoNation, Mitsubishi Sanda, Omron, Panasonic AIS, @seeingmachines & @SmartEyeAB. FN, SM & SE are Tier 2s, all others are Japanese Tier 1s. Japan v.v. competitive for DMS DWs.