Smart Eye has recently made a commissioned study of the projected DMS volumes for 2023E-2026E. The first (base case scenario) was presented in a video published in connection with the rights issue. The second projection, published in the Q4 report, originates from the same study and represents a more conservative scenario. We think this message was not clear to the market, which viewed the latter projections as the final projection.

In a recently published video by Smart Eye, the company provided updated estimates on the number of cars with DMS for 2023 to 2026. The projected DMS volumes were divided between Europe and Rest of World.

For 2026, this video projected 38m cars to be sold with DMS, compared to our previous estimates of 35m.

However, the expected acceleration looked different from what we had previously estimated. In short, we expected a gradual ramp-up, whereas Smart Eye expects initial acceleration to be slower but with much higher acceleration in 2026.

Then, in the Q4 report, Smart Eye provided new projections about the upcoming acceleration. The two projections originate from the same source – a commissioned research study – for which the video presented the base case scenario, and the Q4 report presented a conservative scenario.

The company also presented pre-covid estimates, which showed a higher adoption number for each of the years between 2023-2026 – to illustrate the delays caused by the pandemic and the semiconductor shortage.

Below are the projections presented in the Q4 report (conservative), estimates from the video (base case), and pre-covid estimates:

Breaking down the conservative outlook on a geographical level, the picture is the following:

The assumptions for the above conservative estimates are the following:

As we know that much of the volumes until 2026 are already planned, we believe Smart Eye has very good visibility into these numbers. As a consequence, we will lower our estimates for DMS volumes for 2023-2026.

In Q4 and beginning of Q1, Smart Eye was awarded a big number of design wins – more than doubling its total amount to the current 217. Among these, we are quite certain that Smart Eye managed to ‘steal’ Ford as a customer from its main competitor Seeing Machines.

Ford has so far been the customer Seeing Machines’ has showcased the most, i.e., it has been its flagship customer. Smart Eye winning Ford is thus not just a big win for Smart Eye, but also a major loss for Seeing Machines.

An important detail about this deal is that Smart Eye’s processor partner is Qualcomm – indicating that Seeing Machines’ ‘alliance’ with Qualcomm is weaker than we previously had feared. This is very positive for Smart Eye.

Previously, we considered the alliance between Seeing Machines and Qualcomm as a possible threat to Smart Eye’s market dominance. The rationale behind this was the following:

However, the Ford deal for Smart Eye kills this above thesis, we think. As such, we are becoming much more confident in our estimates of market share for the following years. In fact, we are starting to think they might be a bit conservative.

As we have previously stated, we are quite certain that Smart Eye has won several design wins with General Motors (GM). Among the models is the Cadillac Lyriq, which was showcased during CES 2023 in January. We think the DMS of this model belongs to Smart Eye. Also, it was said that Qualcomm’s Snapdragon processor is the hardware used. As such, we think this is just one of several examples of Smart Eye and Qualcomm teaming up.

On 7 March, Smart Eye announced twelve new design wins from an existing Korean car manufacturer (Hyundai/KIA, we assume). The total estimated value of the design win amounts to SEK350m. Three of the twelve design wins will go into production in Q4 2023, while the remaining will enter production in 2024 and 2025.

We argue that these design wins were very much expected as they are on a platform on which Smart Eye has already been awarded several design wins.

The positive about the design wins is that they will start generating revenues already in 2023. This bodes well for the company’s estimate to become cash flow positive in H2 2024.

For platforms on which Smart Eye has already been awarded a design win, we believe many more similar design wins with a short period until production are likely.

On 27 February, Smart Eye announced its first design wins for medium and heavy trucks ever. The initial order is worth SEK150m with an additional potential of SEK400m. A typical truck is produced over ten years, so this deal spans this period.

The deal is for ‘normal’ DMS (software only) rather than AIS (hardware + software). However, the feature set for these heavier trucks is likely substantially more advanced than a typical car’s. Hence, we estimate the average selling price (ASP) to be at least 50% higher than for a car.

The end customer is said to be a global manufacturer of heavy vehicles headquartered in Europe. The customer is also said to have three 'renowned' brands. Considering that Traton consists of the three 'renowned brands’ MAN, Scania, Volkswagen (and then the less well-known Navistar and Rio), we think the customer is Traton.

Considering that all heavy trucks produced in Europe are also obliged to include DMS by 2026 (and all new models from 2024), we think this is the first among several large tenders for heavy trucks. Apart from Traton, the two major European players to look out for are Volvo and Daimler/Mercedes.

Each year, around four million heavy trucks are produced. Considering that we believe ASP is cEUR10 per truck, we think the TAM is approximately SEK400m per year.

As for personal vehicles, medium-sized and heavy trucks are also forced by law to have DMS in all vehicles produced for the European market from 2026. New models launched in 2024 and 2025 also need to have DMS.

Considering the long planning cycles for the automotive industry and perhaps even longer for heavy trucks, purchasing the necessary technology tends to happen many years in advance. Considering 2024 is only one year out, time is running out for choosing the included technology components.

Particularly, the DMS software needs to be integrated with the adjacent hardware components, not least with the processor. Considering Smart Eye’s and Seeing Machines’ strategies, we think Smart Eye has an advantage in being ‘flexible’. By not being hardware-specific but agnostic, Smart Eye has already made its software ready for relatively fast integration. We think this relatively quick integration with whatever processor is chosen comprises a competitive advantage.

In a recent market update from Smart Eye’s main competitor, Seeing Machines, it stated that it will not participate in RfQs amongst heavy trucks. This leaves this market wide open for Smart Eye.

According to its CEO, the reason why is that the unit economics for this segment is not good enough. This would imply that the higher ASP would not weigh up against the high costs implied.

This can, however, also indicate at least three other things:

First, the set of functions required spans past what Seeing Machines can offer, implying that it would be too expensive to sublicense technologies to be able to compete. This is something that we’ve previously argued regarding the race in Interior Sensing.

Second, Seeing Machines not being able to deliver fast enough. This would stem from not having as flexible software as Smart Eye, combined with Qualcomm’s processor perhaps not gaining as much traction, implying that integration with Seeing Machines would take too long.

Last, and likely a consequence of one or two of the previous points, the major procurements have already been settled on the OEM level, and Seeing Machines not having won any of these. We think this is a likely scenario, considering that Smart Eye would not offer its DMS at a price that was not economically sound. Hence, we would not be surprised if Smart Eye announced two more major truck OEM deals soon.

We have many times stated that being awarded a design win for a new platform is a big deal. The reason why is that this first design win, let’s call it a ‘platform win,' is a strong indicator of the supplier being chosen for all car models on that platform. As a reminder, a car platform is typically the technological foundation for all new cars launched over seven years. Since each car model is generally sold over seven years, a platform win indicates having secured revenues for a full 14 years (seven years to launch of the last car and that being sold for seven years).

In the Q4 conference call, Smart Eye said there is generally no procurement process for follow-up orders. Instead, these follow-up design wins are automatically awarded to the supplier who won the platform. The rationale is that it is very expensive for an OEM to change the existing technology components on a platform. Hence, competitors are invited to compete for follow-up orders only when the OEM sees an obvious flaw in an existing technology component.

As a consequence, the ‘additional potential on existing platforms’ that Smart Eye reports is very probable to be transformed into revenues.

In December, Seeing Machines announced a new automotive program with an existing customer – a large European-based global automotive group. Considering that we believe Seeing Machines has previously been awarded a design win with Chrysler, we believe this is Stellantis. As such, Smart Eye won the first tranche, and Seeing Machines won the second tranche for this customer. This was expected considering that we believe all the major OEMs like VW Group, Stellantis, and Toyota will use dual or triple sourcing.

On 14 March, Cipia announced a design win for driver and cabin monitoring system for an American electrical vehicle manufacturer. Considering that Tesla makes its own DMS and OMS, this is likely one of the smaller ones like Rivian or Lucid. In 2019, Smart Eye took a strategic decision to only focus on large OEMs. As such, Smart Eye did not participate in this procurement.

With the recent major design wins, Smart Eye now has a total of 217 design wins. The value of awarded design wins now amounts to SEK4.4bn, and the value of additional models on already won platforms is SEK4.1bn. Hence, the total value of “probable orders” is now SEK8.5bn. This is distributed between 19 OEMs and 18 platforms.

As usual, it is difficult to compare market leadership on an apples-to-apples basis. However, we currently estimate Smart Eye to have won 50% of all design wins, but the market share for the underlying order value is even higher (due to Smart Eye and Seeing Machines selling at a price premium).

Looking at Smart Eye and Seeing Machines only, we have made the following table:

In the conference call, the company was asked about its competitive position in interior sensing. Following the acquisition of Affectiva in 2021, Smart Eye claimed to have gained a competitive advantage in interior sensing in terms of now having acquired a bigger share of the functionalities on OEM’s checklists.

On the Q4 conference call, the company reiterated that it is still “years ahead of competition in interior sensing.” The main headstart still stems from the technical features acquired through Affectiva, meaning the ability to sense a person’s experience through Emotion AI.

Smart Eye stated that it showcased these features at this year’s CES in Las Vegas and that it now has several discussions with OEMs (likely the European premium ones like Mercedes, BMW, Audi, etc.) about including these in their cars.

The last major DMS procurement is Toyota – the world’s biggest car manufacturer in terms of produced cars (c11m per year). On the conference call, Smart Eye stated that Toyota would be more fragmented in its sourcing than other large OEMs. Smart Eye expects Toyota to do one procurement for each car platform, i.e., using a multi-sourcing approach.

Looking at the upcoming car platforms of Toyota, we have the following:

All in all, there are four platforms, of which three are new. Hence, we guess that three separate procurements will be completed.

Smart Eye announced its first order for its aftermarket product, AIS, amounting to SEK40m but SEK70m for expected follow-up orders. The first revenue from this order is expected in mid-2024. In our opinion, this is a ‘free optionality’ in the Smart Eye case. A bit surprising is that the deal was for a European luxury sports car manufacturer (Ferrari?) and not for the commercial vehicles segment.

However, considering that Smart Eye does not typically participate in RfQs for very small OEMs, and that this type of OEM typically differentiates itself in ways other than through safety features, we think this was a last-minute solution to meet the 2024 regulations.

In the final words from the Q&A session in the Q4 conference call, Smart Eye made the following outlook statement for 2023:

This implies at least 95 design wins. We expect several more to come in the heavy-trucks segment, perhaps one chunk from Toyota, and then several follow-up orders from platforms on which Smart Eye has already been awarded one or more design wins.

2023 should be the year when design wins are starting to be awarded for interior sensing/cabin monitoring. While there have been previous claims from some competitors about cabin monitoring design wins, we think these have all been “DMS+”. Smart Eye expects these design wins to go into production in 2026 or 2027.

Considering that the only announced order within AIS is expected to launch in mid-2024, this was a positive surprise for us. Considering that this is an aftermarket product, it could be the case that large orders do not exist, but rather small continuous orders being executed from several fleet customers. For 2023, we estimate revenues from AIS of only SEK28m, implying a considerable upside potential to a segment expected to ‘ramp up properly’.

No surprise but an add-on to the statement from the video from February about estimating accelerated growth in Human Research in the US in 2023. Considering the 15% organic growth in this segment in 2022, we estimate this ‘accelerated growth’ to imply c20% y/y growth.

Considering the cost-cutting program announced in December, but also the rapid ramp-up of the cost base before that, we think it’s worth to re-calibrate what the underlying cost base is. This is particularly important when assessing the road towards Smart Eye expecting to become cash flow positive in H2 2024. This is how we reason:

In Q3, reported OPEX incl D&A was SEK139.6m. Of this, D&A amounted to SEK39.5m, implying that OPEX was SEK100.1. Investments in intangibles amounted to SEK22m, implying that the total cost base is SEK122m per quarter, or SEK488m per year.

Then came the announcement of a cost-savings program, in which Smart Eye claimed that it would reduce its personnel costs by cSEK30m/year. In the Q4 conference call, management reiterated that most of the measures were implemented already in Q4 2022. Then, the remaining cost cuts will be completed during Q1. No one-off costs are expected to incur as the cost base consisted of consultants.

By taking the SEK488m minus SEK30m, we get SEK458m on an annual basis, or SEK114.5m per quarter. We then estimate a 5% increase of the cost base in 2024.

To become cashflow positive in H2 2024, Smart Eye’s management and board of directors have implemented stricter quarterly budget meeting to make sure the company won’t require additional cash.

Management also states that it is committed to reach its target – and stated in the Q4 conference call that it will cut more costs if needed to become cash flow positive in H2 2024.

From Q4 2022 to Q4 2024, we estimate Smart Eye to increase quarterly revenue from SEK55m to cSEK140m. The main driver is the ramp-up of DMS license revenues. For license revenues, we use Smart Eye’s recently provided conservative projection of DMS adoption.

We also estimate a c15% increase in quarterly NRE revenue from today’s levels. Additionally, we estimate c20% annual growth in Human Research, backed up by Smart Eye saying in a recent video that it expects accelerated growth for this area in 2023.

We estimate Smart Eye reaching breakeven on a cash flow basis in November or December 2024. In this scenario, we estimate the company to have cSEK43m to spare before starting to generate own cash flow. Should the company need it, Smart Eye has two additional unutilized bank facilities of SEK5m and DKK8m (SEK12m), bringing the total estimated headroom to SEK60m.

We think Smart Eye appeared quite confident in reaching its target of becoming cash flow positive in H2 2024. Considering that we estimate this to be completed by a somewhat narrow margin, the certainty from Smart Eye may seem strange.

In the conference call, however, Smart Eye stated explicitly that it has continuous dialogues with all its OEM customers about expected start of production (SoP) dates, as well as expected ramp-up volumes. As such, the company is likely to be able to make quite good projections in the short term.

While we think there is some headroom before running out of cash, the distance to zero is not enormous. Should there be a risk of more cash needed, we think a short-term credit line would be possible to issue. However, we think and hope that Smart Eye would cut costs before such an event would happen.