Image source: Bosch

Mobileye’s main strength in ADAS in incumbency, but it looks determined to unite almost the entire automotive ecosystem against it. The closest historical precedent in the tech space is probably Intel uniting the embedded processing industry around Arm to compete against the X86 architecture, including two obscure, niche, suppliers called Nvidia and Qualcomm.

Back at the start of the century, Intel was far bigger than any one Arm competitor. But by pulling the ecosystem together, Arm steadily created a force Intel could not compete with. Each time any Arm licensee won against Intel it eroded Intel’s share.

For wont of a better description, we could call it Patriot missiles vs. a drone swarm.

Intel had to win everywhere to maintain share. Arm won each time a licensee won. What did Arm do? It created a portfolio of IP that covered the entire embedded processing waterfront. As I wrote many times in my then BROADCAST newsletter, “From M0 to A57: Sensors to Servers.”

In automotive, Qualcomm is broadly following the Arm playbook. First in the cockpit, then in ADAS, most recently in mixed-criticality SoCs covering ADAS and cockpit. So Mobileye is taking the Intel approach and doing everything itself, while Qualcomm is building an ecosystem and partnering with practically everyone.

In ADAS we can already see Qualcomm has three routes to market:

• Qualcomm Snapdragon Ride Pilot: developed in partnership with BMW.

• Qualcomm partners with Wayve.

• Qualcomm partners with Bosch.

There will be others, but we will look at those as they come to light.

Qualcomm took about three years to convert the Arriver driving stack acquired from Veoneer into the Ride Pilot stack developed jointly with BMW.

BMW was a key Mobileye partner for many years, using its EyeQ platform up until the iX model launched the iNEXT platform in 2021. But the relationship either soured or floundered, and BMW dramatically pivoted to Qualcomm around 2022. Ride Pilot now powers the BMW Neue Klasse platform, including the MY26 iX3 and i3 models.

Likewise, Bosch has quietly transferred its suite of proprietary ADAS and automated driving software to run on Qualcomm Snapdragon Ride SoCs and Snapdragon Ride Flex. As the new kids on the block, the partnership with Wayve is harder to judge, but we will know more in time.

Over a long enough period of time, water can erode rock. Just as with Arm, Qualcomm doesn’t care who wins. Any win for a Qualcomm SoC partner is an erosion of Mobileye’s market share in ADAS. Such is the nature of a war of attrition.

Mobileye looks like it is doing everything it can to accelerate this attrition. Let’s look at tier 1 suppliers and driver monitoring systems for more clues.

Read more

Mobileye highlights a key point in its PR, that: “these programs reflect growing OEM demand for the consolidation of driver monitoring, occupant safety, and advanced driving functions – eliminating the cost and complexity of a separate DMS ECU.”

In automotive, the battle ahead is firmly tanks vs. bayonets. Let’s look at the details.

Source: Mobileye

Read more

Source: BMW

As the press images show, the interior of the i3 is broadly the same as that revealed in the iX3, and for the driver monitoring industry, the i3 also has combined driver and occupant monitoring located in the rear view mirror chin. For now the assumption is the new platform will move the driver monitoring and cabin monitoring features into the mirror location in all Neue Klasse models.

With the passing of time we slowly get more data points about the apparent procurement decisions made by BMW for Neue Klasse. The iX3 design is discussed below.

Let’s have a quick update here with what else we may be able to unpick.

Read more

Driver monitoring approaches aim to infer driver state based on behavior, typically through camera-based monitoring and vehicle inputs (e.g., vehicle lane position variability). While camera-based driver monitoring systems (DMS) are becoming more prevalent in advanced driver assistance systems, NHTSA has found most of these systems are intended to detect driver drowsiness, inattention, and sudden sickness. DMS-based approaches aim to detect indicators of driver alcohol impairment (such as eye glances, facial features, and vehicle kinematic metrics) and potentially provide some indication of the level of impairment. NHTSA will research characteristics of DMS-based impairment approaches as they are made commercially available and, more recently, has identified one announcement of an effort to introduce DMS for use in production vehicles.

Source: Smart Start

In early February Smart Eye declared total victory in impaired driving technology leadership with the acquisition of startup Sightic Analytics. The analysis can be read here:

Let’s get in to the meaning of the NHTSA update as it relates to impaired driving legislation.

Read more

In its latest quarterly results published earlier in February for the periods 4Q25 and full year 2025, Smart Eye management reiterated once more that statements of quarterly cars on road, software pricing guidance, and segmentation of the automotive revenue total into automotive licensing, NRE and aftermarket AIS would not be provided.

The analysis is provided here:

Seeing Machines has now released a presentation which shows all this data. Let’s take a look.

In a rough calculation, automotive royalties was about 80 percent of first half FY2026 automotive revenues, which covers the period July to December. If total automotive revenues in that period is about US$12 million, that makes royalties about US$9.6 million. This is really broad-brush analysis, but it serves the purpose.

Delivery data from Seeing Machines is stated quarterly and released into the public domain. Deliveries for the second half of 2025 totalled 1,088,530 vehicles.

ASP = revenues/units.

So the Seeing Machines ASP for its software license is about US$9 over that period.

See how easy this is? Did the sky fall? No.

The estimate for Smart Eye is more like US$3-5 per vehicle for DMS. As is always the case in life, you get what you pay for. You can read more here:

July 2026 sees the introduction in Europe of the General Safety Regulation, as well as the introduction of the 2026 testing regime for Euro NCAP. Long story short, in third quarter 2026, every vehicle with four or more wheels sold in Europe (including EU, EFTA and the UK) requires a vision-based DMS to monitor for advanced distraction.

Data from the ACEA suggests new registrations of about 13 million light vehicles in 2025. First thoughts here:

More broad-brush numbers: that means in the period July to September with GSR fully in effect that between 3.0-3.5 million light vehicles will be registered in Europe with a DMS.

What is the Seeing Machines DMS delivery estimate for the coming quarters? That is shown below. July to September equates to FY27Q1. Note these are global delivery numbers.

What the chart demonstrates clearly is the role of EU regulation to drive up DMS volumes. What is coming this year will be dramatic, with volume growth starting in earnest probably in the January to March quarter. Which is now.

Does Seeing Machines know precisely the total volume shipments in each of the coming quarters? No. But neither do the tier-1 suppliers or even the OEMs. Those numbers are only known for certain after the period has closed and the hard data has been analyzed.

So here is the issue: Seeing Machines puts its delivery numbers and forecasts into the public domain. It publishes segmentation of its automotive revenues in sufficient detail to perform meaningful analysis, even to the point of calculating the average software price.

Smart Eye simply won’t do it.

Decide for yourself who is actually leading the DMS market as we go into the rapid growth quarters ahead for GSR compliance.

Want to know more about the big picture for the European automotive DMS market and "who supplies whom?" Message me for details of the Semicast analysis: "2026 Forward Share Estimates of the European Automotive Driver Monitoring Systems Market." There are two reports, one covering Tier 1 suppliers, and one covering Tier 2s.

Act now to request your copy of the study brochures and order forms to purchase these reports and to cut through the industry PR noise. Fastest delivery is by credit card payment via Stripe, accepting various card types. Direct wire transfer and conventional P.O. routes also accepted.

Subscribe now

© 2026 Semicast Research Limited. Do not reproduce.

Below are my first thoughts on the numbers:

Read more

Market analysts, financial analysts and institutional investors need data to make informed decisions and analysis. We don’t need Daddy to tell us everything is wonderful. We are adults and judge the facts for ourselves.

Nothing about Smart Eye makes sense, because we are intentionally being kept in the dark. You will have to decide for yourself if that constitutes deceit, but based on the market’s current reaction to the 4Q results and live Q&A, a swathe of investors have simply had enough and just said “stop.”

The commentary from CEO Martin Krantz is inconsistent. In the 3Q25 Q&A in November, license growth was happening across the board. Fast-forward to the 4Q report today and he says: “The license revenues are growing. In this quarter the European deliveries jumped rapidly, almost doubling from previous quarter, while the license revenues from the rest of the world came in softer.”

So, Europe up, but everywhere else suddenly softer?

The European numbers should be growing in the run up to GSR. The issue isn’t that they are going up, but that management continues to insult the intelligence of everyone by withholding the raw data.

“Almost doubling” is equally true of delivery growth from 25KU to 49KU as it is from 250KU to 495KU. But the implications for market share and cash generation to reach profitability are vastly different. We can’t judge the true picture, because the company won’t tell us.

Here’s another example from the Krantz commentary, this time on intoxication:

“By mastering the trinity of distraction, drowsiness and intoxication we are leapfrogging the rest of the DMS market. We have developed this feature for over five years and started to market this feature last year and in early 2026 we both received our first design win and decided to partake in the ongoing market consolidation by merging the market leader Sightic into Smart Eye. This move enables us to further accelerate the development and enables faster refinement of algorithms, giving us a unique positional advantage on the DMS market, ultimately leading to many more lives being saved.”

Krantz then said in the Q&A that he hasn’t seen the Seeing Machines intoxication technology, doesn’t really understand the regulatory requirements for intoxication detection in the US, and is developing technology which isn’t really aligned with the NHTSA feature roadmap.

But don’t worry about any of that, because Krantz tells us Sightic is the market leader and Smart Eye/Sightic together magically leapfrogged the rest of the DMS market.

This wasn’t an investor presentation. It was Daddy telling a bedtime story. But the monsters are lurking under the bed.

Let’s unpack three issues:

1. “Design wins” as proxy for market share.

2. Smart Eye market share forecasts.

3. Automotive revenue split between NRE, software licenses and AIS.

Read more

My first thoughts are as follows:

Read more

Read more

Fine. Let’s do our best to fill in those blanks.

The commercial viability of the three publicly-traded driver monitoring systems software companies - Seeing Machines, Smart Eye, and Tobii - is defined by the successful balancing of the following three priorities:

1. Don’t run out of cash.

2. Continue to invest in a technology and innovation roadmap to remain competitive.

3. Achieve profitability to create a long-term sustainable business.

A company does not die because it makes a loss, it dies because it runs out of cash. So any supplier failing the first priority is dead. But priorities (2) and (3) are contradictory: In driver monitoring, cutting costs just to keep the lights on kills the long-term technology and innovation roadmap needed to remain competitive.

Cut too much and short-term survival simply translates to long-term death.

Cuts, cuts, more cuts

Read more

For now we can set aside the large DMS market in China, which has different software suppliers, different algorithm requirements, and wildly different price points. That is not to say the Chinese DMS algorithm developers are not succeeding in Europe - they are. But Europe is raising the bar for driver and occupant monitoring system performance through Euro NCAP, now with the introduction of NCAP26 and next with NCAP29.

Euro NCAP is vitally important because its test protocols tend to be replicated by other regional NCAPs around the world. ANCAP (Australasia) and BNCAP (India) are both following the Euro NCAP roadmap; JNCAP (Japan), KNCAP (Korea) are poised to follow Europe. CNCAP (China) is following its own path and timescales. Even NCAP (the original New Car Assessment Program which covers the U.S.) is stirring into life for DMS and OMS for hands-free driving assist and monitoring for distraction, drowsiness and impaired driving.

So while much commentary on DMS is focused on EU regulations, it is actually the voluntary adoption by OEMs of DMS and OMS to maximize points scoring for NCAP that is driving volumes in many regions.

That is why what is happening in the European DMS market is so vital to understand, because Euro NCAP sets the tone for what happens in much of the rest of the world.

For DMS in Europe, pole position comes down to a straight race between Seeing Machines and Smart Eye. Other algorithm developers are active, but they have a small estimated market share compared to these two suppliers. That may change in the coming years, but as the facts change, the analysis will change with it.

For 2026, this is a two-horse race. Let’s dive in.

Read more

2025 was a wild ride for the driver monitoring systems industry. Cipia’s automotive assets were swallowed by Harman, with Tobii Autosense closing the year with a life-saving licensing agreement with an unnamed automotive supplier and a new CEO. In order to stay independent, Seeing Machines reduced headcount, with Smart Eye closing new debt-based funding. Last year also saw a strategic coming together of indie Semiconductor and emotion 3D, with Mobileye blogging about a proprietary DMS running natively on its EyeQ ADAS processors - a solution it has yet to show in public.

Developments in the industry last year were mostly the result of a universal decision among OEMs to adopt steering wheel torque sensor solutions to meet compliance with the EU’s General Safety Regulation (GSR) for drowsiness detection. GSR forces OEMs to adopt vision-based solutions for advanced distraction warning, but that comes into force in mid-2026, thus buying the OEMs a 24-month window to delay installing mandated life-saving technology.

In automotive, saving lives always comes second to saving dollars, and that 24-month delay spelt the end of the road for some DMS developers, and a rapidly revised strategic focus for others.

The DNA of any company is exposed when it is faced with an existential threat. For some, the response is helplessness and an inevitable slow march to bankruptcy as talent leaves. For others, it is a call to out-innovate and out-think the prevailing circumstances. At CES we saw the first example of out-innovating and out-thinking the situation, with a technology pivot from Seeing Machines away from traditional driver monitoring and into interior perception mapping (IPM).

Read more

Read more

All of these threads are suddenly coming together to bring about significant new opportunities for distraction and drowsiness monitoring in big rigs in the U.S.

Source: Peterbilt

On November 12, Seeing Machines announced an initial order for 1,100 Guardian Gen. 3 aftermarket units across the distribution fleet of a major U.S.-based multinational company. With installation to be completed by December, the customer evidently wants part of its distribution fleet protected ahead of the busy holiday season.

The identity of the customer in unknown, but obvious possibilities include Amazon, Coca-Cola, PepsiCo or Walmart. The actual name doesn’t matter. What matters is that a company with a large truck fleet has undertaken an exhaustive field trial of all suitable solutions and then selected the optimal drowsiness mitigation technology.

Read more

Source: Sense Media / InCabin

Details of the Magna occupant monitoring mirror can be read in the linked post below:

Also presenting its driver and occupant monitoring mirror at InCabin was Gentex, and shown below is Brian Brackenberry on stage at the event.

Source: Sense Media / InCabin

Details of the Gentex offering can be read in the linked post below:

Let’s look at the new development in more detail.

Read more

As always in automotive, news headlines are made and nothing happens for years. Then Qualcomm and BMW stole the show at IAA with Ride Pilot. The resulting software feature set, shown below, is extensive:

Source: Qualcomm

Having witnessed nothing tangible from Qualcomm on the AD software stack since 2021, Ride Pilot is Qualcomm’s inevitable entry into the automotive drive software market. Under the terms of the agreement, the stack has become the property of Qualcomm, which can now license this software to other automakers directly (on a “software tier 1” basis) or via traditional tier 1 suppliers such as Valeo.

Valeo and Qualcomm announced details of their extended technology partnership on the opening day of IAA:

Analysis of the announcement suggests Valeo is building its own platform for both ADAS and AD applications based on the Qualcomm software stack. The key software components of the Valeo offering look to be:

• Ride Pilot software stack.

• Valeo parking stack.

• Seeing Machines DMS/OMS stack.

Let’s dig in further.

Read more

Demonstrated on the Magna stand at the Munich IAA Mobility show in September was a transparent version of the mirror, as shown below.

Source: Colin Barnden

Clearly visible in the pictures is the CMOS image sensor in the center of the module, and on the right hand side a horizontal set of IR illuminators and two vertically mounted pairs of illuminators.

The orientation and position of the illuminators creates a highly distinctive flash pattern, making the Magna mirror easily recognizable from YouTube videos posted by influencers doing test drives of new model releases.

940nm IR light is invisible to the human eye, but is detected by the CMOS image sensor of most modern digital cameras and smartphones, creating the now distinctive purple hue seen in still images and full motion video.

Also revealed for the first time at Munich IAA was the second generation of the Magna occupant mirror, as shown below. Let’s unpack that in more detail.

Read more

Image source: Colin Barnden

As part of the presentation, Duggal provided details of the ecosystem and partners used to develop the Ride Pilot software stack in the iX3, as shown below:

Read more

Source: Smart Eye

The latest Smart Eye analysis shows a 2026 global market for DMS estimated at about 20 million units, with Europe dominating shipments at about 12-13 million. While the specific numbers are open to debate, the importance of success in Europe isn't.

ADDW & NCAP

Mandatory ADDW compliance for GSR equates to an incoming 100 percent fitment rate for DMS in light vehicles sold in Europe. Meanwhile the Euro NCAP 2026 test protocols force OEMs to shun cheap and poorly performing DMS for more sophisticated systems which clearly score above the rising minimum threshold for a 5-star rating.

For NCAP23 OEMs could (and did) mostly ignore DMS features and performance and focus instead on scoring points in other categories. But for NCAP26 a DMS which fails to score above the minimum threshold in the “Safe Driving” category will automatically disqualify the entire vehicle from receiving the coveted 5-star safety rating.

The combination of these two factors both accelerates uptake of DMS in Europe over the next twelve months, and directs OEMs to just a handful of automotive DMS algorithm developers (Tier 2 suppliers) with a suitable feature set and advanced technology roadmap that can meet the necessary performance requirements.

Most automotive OEMs typically work on two-to-four-year design cycles, with Chinese OEMs typically working on 12-month cycles. This means the procurement decisions for which DMS algorithm developers will be used for ADDW compliance and NCAP26 in 2026 have already been made by every light vehicle OEM selling into Europe.

Read more

Source: YouTube

Revisions to the conclusions based on additional research now point to the DMS developers shown here being Seeing Machines on the left and Asaphus Vision on the right. Seeing Machines acquired Asaphus from Valeo in July 2024 and Valeo is assumed to be the tier 1 supplier of both systems shown.

No visual evidence of Smart Eye DMS in any Stellantis model at start of production has so far been found, but this may change in the coming quarter as every OEM races to bring models to production featuring DMS ready for the coming changes in 2026.

Read more