Information & Communication Technologies


CES 2017 Automotive Impressions – Frost & Sullivan Viewpoint

by Praveen Chandrasekar 16 Jan 2017
Share this:

Artificial Intelligence in Different Sizes and Digital Assistants Steal the Show

Many including the team at Frost & Sullivan initially believed that CES 2017 was a bit of a letdown compared to 2016 but on a closer look it actually made more sense on a few focused areas. Though there were a few important misses like GM who chose to stay out of the show this time, many including the likes of Volkswagen did not present anything new except for the Amazon Alexa integration. The real action of the show from an automotive angle was actually happening in the corners of the North Hall where companies such as NVIDIA and Qualcomm were demonstrating.  

And interestingly even though automated driving was still the star of the show there were many showcases that did not simply just focus on the core functionality but focused more on areas like building trust with the user and applying different levels of artificial intelligence (AI) to support the driver in many ways. This blog is not about going into each and every OEM or supplier or silicon vendor or startup and detailing what they showcased but more a tool to capture what key trends and themes emerged from the show according to Frost & Sullivan.

Frost & Sullivan observes the following as key themes and trends emerging from CES 2017 from an automotive perspective.

  • Artificial Intelligence is Huge but it comes in Different Shapes and Sizes
  • Speech is becoming an Uber Cool Input HMI Option Again.
  • Startups Positioning BEV’s in the Ultra Luxury Market
  • LIDAR Game Slowing Down & Left with a Few Pragmatic Players
  • User Personalization and the Role of Biometrics
  • Cybersecurity Moving into the Silicon Zone
  • HD Mapping is becoming More Competition Intense but HERE Gaining Traction
  • Android N is not Any Close to OEM Production Ready
  • Answers Emerging for Helping OEMs Monetize Vehicle & User Data
  • Tier 1 Suppliers Finally Embracing Service/Solution Business
  • HMI Innovations similar to last year

Artificial Intelligence is Huge but it comes in Different Shapes and Sizes

Interestingly many claim NVIDIA as the star of CES 2017 from an automotive angle and attributing this purely to their AI announcements and partnerships revealed at the show. The long list of collaborations showcased at CES includes Audi, Daimler, Bosch, ZF, HERE, Zenrin, TomTom, Elektrobit, Baidu, etc. All or most of these utilize the Drive PX2 as the core of their AI initiatives focused mostly on self-driving cars except for ZF which is taking this technologies into Factories in 2018.  But most importantly this is not all the AI we got to see at the show, whilst NVIDIA’s work is centered around AI for actuating the driving experience by a deep sense of understanding of the environment, there were other AI showcases that did not get into the actual driving and were more focused around bring contextual awareness and driver understanding and baking them into digital speech assistants in the car. While the later on digital speech assistants is a growing trend where Amazon Alexa dominated the show not just from an automotive angle, the former around AI for contextual awareness was very nicely demonstrated by IBM Watson and Local Motors team on an autonomous shuttle and the concept really stresses on building trust with the user.

Frost & Sullivan recently completed a study on self-learning cars and this segmentation of AI from a feature level (e.g. Voice), Contextual Level (e.g. Driver Understanding) and actual driving level that we showed in the report is what we saw in many ways at the show. AI for autonomous driving involves hardware and neural networks, learning platform which makes NVIDIA almost a tier 0.5 in this equation whereas AI contextual awareness involves pure software which is where IBM is playing with Watson.  Automotive speech companies like Nuance are integrating AI into their Digital Assistant (Available on select 2016 and 2017 BMW models called as Dragon Drive).  Hyundai showcased AI from a wellness perspective and Toyota's take was more about vehicle and user building trust (the most critical one for autonomous driving). On the whole while we will not witness AI in autonomous driving before 2020 we will definitely witness AI and deep learning from a contextual perspective in a NVIDIA supported Mercedes Benz car in a year’s time.

Speech is becoming an Uber Cool Input HMI Option Again

Two vehicle manufacturers namely Ford and VW showcased integration of Amazon Alexa, Nissan showcased Cortana integration, FCA & Daimler showcased Google Assist and the common theme is all of these are consumer level digital speech assistants that are fast finding their way into the car from a vehicle to home, home to vehicle (in Ford’s own words) and in-car experience perspective. Amazon Alexa might just be in 12 million plus devices but with a smartphone integration deal announced at CES (Huawei) and couple of vehicle manufacturer partnerships might just take this free to use solution to a next level. And more importantly given that SIRI and Google Assist were slow to open up to third party developers and Alexa from the beginning hit on that point might make this more suitable to the automotive environment. But all of this integration will be useless if the reliability and lag issues still plague the speech market in the car-cabin. The focus is coming back on hardware as against software and cloud-enabled VR which are still lagging in quality because of connectivity issues in the car. And interestingly a startup named VocalZoom based in Israel showcased this at their booth in Honda with a new optical sensor that can accurately understand your speech by understanding gaze/facial muscle movements. The demonstration was amazing and its concepts like these that would enable OEMs to transition voice as the primary UX weapon because while everything else is physically distracting and voice is only cognitive distraction, this can be solved if you have intelligent digital assistants that can prompt instead of making you think. And once the integration starts happening across several cars then AI kicks into this area as well where user understanding and personalization can happen. Frost & Sullivan’s recent research around voice recognition in the car covers many of these interesting trends and technologies and even areas like voice biometrics.

Startups Positioning BEV’s in the Ultra Luxury Market

One could be a skeptic of how successful Faraday Future or Karma Automotive or Lucid Motors or NextEV will be but as they claim since they are starting from a blank sheet of paper they have much lower challenges except for building a brand and remaining cash liquid. But the big takeaway from the show was how the German luxury makers missed the BEV game big time but are catching up now with OEMS like Daimler going to pump 10$ billion into BEV's and everyone from BMW to Volvo gearing up big time to aggressively penetrate this market. This theme was also evident from the recently concluded NAIAS show in Detroit. Frost & Sullivan’s analysis of the Luxury market and BEV luxury market and benchmarking of the different OEMs clearly indicate that Tesla is in  a pole position in the BEV luxury market and most of the new entrants are taking their model and applying finer enhancements to range, performance, technology and most importantly interior look and feel.


LIDAR Game Slowing Down & Left with a Few Pragmatic Players

This year also witnessed startups like Innoviz (At the Magna booth) promising a $100 solid state 200M range LIDAR in a couple of years. Leading tier 1 supplier Continental announced a 3D flash LIDAR that will have a range of 200 m that will be in production by 2020. Quebec-based LeddarTech unveiled their new line up of 2D and 3D LIDAR with range of close to 200 m. Despite all these announcements around solid state LIDAR, the reality is that suppliers like Valeo and Velodyne who supply mechanical LIDAR currently are going to benefit more in the short term with actual production contracts. While LIDAR slowed down suppliers like Autoliv who are transitioning very fast from pure occupant protection to accident prevention and automated driving presented a range of solutions that utilize sensors that are available today such as mono vision camera, stereo camera, Long range and short range radar, driver monitoring camera, etc. What was amazing about Autoliv’s demonstration besides the industry leading night vision system with emergency braking to autonomous emergency braking using a mono-vision camera was the LIV research vehicle that focused on the sharing between the vehicle and the driver in autonomous mode using existing sensors and a central ECU, this concept also involves artificial intelligence. Frost & Sullivan’s recent research around the LIDAR market clearly indicated that some of the early birds in the solid state market are actually struggling to deliver products to the market on challenges relating to manufacturing while there is also a view among many OEMs to make automated driving work without LIDAR like Tesla.

User Personalization and the Role of Biometrics

With automated driving and alternative mobility gaining traction, one area that becomes interesting and valid is user-based personalization across cars including interiors, radio stations, etc. Continental demonstrated this using biometrics in combination with a smartphone key on the joint venture OTA Keys but other OEMs are simply doing this using a facial recognition camera ( and or adding biometrics as a two factor authentication). Remember this camera that serves as user identification can also serve as a driver distraction monitor and driver awareness monitor that can be utilized in the event of a switchover between car and user. This is one of the areas that NIVIDA and Mercedes Benz showcased on the Co-pilot application which is adding AI to the driver monitoring by including the environment perspective. It is very clear that the vehicle access market is going to go through a rapid transformation in the coming years beyond passive keyless and smartphone key into one that utilizes technologies like biometrics and facial recognition to allow users to carry their profile across cars. Frost & Sullivan has recently published a study around vehicle access systems that covers a lot of these trends ranging from biometrics to facial recognition to use cases in the shared mobility market.

Cybersecurity still Doesn’t Garner Appeal

Post the acquisition and integration of Towersec, Harman showcased an enhanced version of the ECU and TCU shield at their booth adding to the 5+1 security framework. This year and 2020 will witness Harman announcing customers for both the minimal Towersec solution and the full blown 5+1 framework. Interestingly Argus Cybersecurity who one can claim to be the largest independent automotive security vendor currently was present across many booths including the likes of Qualcomm, Elektrobit, etc. The solution they showcased at the Qualcomm booth was a Snapdragon 820A processor integrated connectivity protection suite aimed at protecting the infotainment and telematics functionalities. Despite these companies and their showcases and others like Iredto and Karmaba also joining the fray, the fact remains that Cybersecurity is still not a consensus topic among OEMs. Frost & Sullivan released a comprehensive study in early 2016 that talks about these companies and other startups and critical technologies that are emerging in this space.

HD Mapping is becoming More Competition Intense but HERE Gaining Traction

Similar to LIDAR, HD maps is another space that is getting increasingly crowded with newer startups like Civil Maps and traditional suppliers like Mitsubishi Electric joining the fray. Civil maps showcased their HD maps concept and also announced a partnership with Quanergy to use their LIDAR for Civil Map’s Atlas DevKit – essentially a data collection system. Mitsubishi Electric on the other hand showed a HD map that they have started developing in Japan using a local consortium there and also showcased a concept vehicle with all the sensors hooked in for the data collection. While all of this is good development in the market, the real development is the direction HERE is taking with a brand new team at the C-level and immediately below and investment from different directions like Intel and Tencent/NavInfo allowing them to move into consumer and Chinese market. While it is clear that company might emerge as the defacto standard in the market with even players like Mobileye announcing partnerships with them for their REM product, the interesting part is to see how they will fare in markets like Drones. One area that is increasingly getting clear is that HD maps will be far more crucial in the SAE Level 4 and 5 automated driving because of the sheer number of variables and challenges on the road and HERE also has a partnership with NVIDIA to utilize their AI for this. Frost & Sullivan’s recent study around HD maps for automated driving covers all of these players and challenges and current market development and where vehicle manufacturers are in this specific area on partnerships.

 Android N is not Any Close to OEM Production Ready

FCA created a lot of buzz when they and Panasonic Automotive made a lot of noise around the Android N integration as a future Uconnect system. While the integration which utilizes the full suite of Android N including Google Assist integration was very neat on a 8.4 inch capacitive touch screen, the disappointment was the fact that the system is nowhere close to commercial production. Partners such as Panasonic Automotive and QNX also made noise around this system that for once was a truly connected experience in an embedded setup with the phone nowhere in sight and allowing the full suite of Google location-based services to be used. Current market leaders like QNX chose to keep it safe with their announcements around automated driving with Renesas Electronics, University of Waterloo, Polysync and use of hypervisor technologies. And GENIVI was also making a lot of noise in Bellagio across a number of different vendors. The conclusion here is that OEMs seem to be in favor of Google under the dash as well but are not quite ready to commit to them and Android N might be the silver bullet that Google was hoping for. Frost & Sullivan is currently supporting a number of vehicle manufacturers on their future IVI OS strategy.

Answers Emerging for Helping OEMs Monetize Vehicle & User Data

This has been a pain point or a challenge for vehicle manufacturers ever since the concept of big data caught steam in the automotive market. While it is clear that vehicle, user and a plethora of other data has several use cases what was not clear was the consumer willingness to pay, privacy concerns and how automakers can actually pull this off internally. Even though not demonstrating, one startup that the Frost & Sullivan team met and was amazingly impressed was Otonomo, an Israeli company that is looking to solve this data monetization problem for automakers. With a strong bunch of investors, the company is creating a connected car value exchange and marketplace that utilizes the car data through an integrated Otonomo API that adheres to the strictest of privacy and security standards. This data is then used by the company to power a number of use cases ranging from insurance to safety to predictive maintenance, retail, etc. where they have established an ecosystem of players through whom this data can be monetized in terms of services that can be offered to drivers, cities, etc. This in many ways is a simple yet revolutionary way to monetize car data and keep out data hungry players like Google out of this space. Frost & Sullivan will very soon be embarking on interesting industry and consumer studies focused on data enabled use cases and monetization strategies where all of these will be covered in detail.

Tier 1 Suppliers Finally Embracing Service/Solution Business

A challenge that remains for tier 1 suppliers is to create a sustainable high margin service business that offshoots their loses from a low margin component business. While many suppliers realize the importance of this none of them have acted in disruptive ways thus far. Two suppliers that are increasingly getting close to this are Bosch and Continental. Bosch was primarily using its IOT business in play in the mobility scenario presenting a range of concepts namely connected horizon for predictive driving in the navigation context, connected parking that utilize a new set of skills in the company. Similarly Continental showcased advancements in its eHorizon framework and Road Database offerings. While these are nowhere close to a full blown service business it is atleast a step in the right direction for these suppliers who are now challenged by not only technology players but interesting startups who automakers are not shy of working directly with. Frost & Sullivan’s recently concluded emerging startups in the automotive and mobility market globally covers a lot of these interesting startups that the suppliers and automakers need to watch out for.

HMI Innovations Similar to Last Year

Haptic feedback was a common theme for HMI this year at CES 2017. Bosch, Denso and BMW all showcased an innovative gesture based haptic feedback system in thin air. What this really means is that the control of the IVI screen happens through air gestures but in this case the difference is the haptic feedback the driver receives once the command is executed. BMW showcased this as an evolution to the air gesture system that is available on the 7 Series and dubbed it HoloActive Touch HMI. Denso showcased a similar system on a holographic virtual touch screen that provided haptic feedback in mid-air. While Denso was quick to say that this system is more applicable to the user in the automated driving mode the others simply showed this off as a next generation innovation. With air gestures themselves being considered as a fancy limited value gimmick adding haptic feedback to it is in Frost & Sullivan’s view pushing it further out on the production timeline. Besides haptic feedback, other interesting HMI concepts were the production ready DMD Head Up Display (HUD) from Continental, futuristic HUD in partnership with Digilens from Continental that can do AR, FlexConnect infotainment system from Mitsubishi Electric supporting three screens running Android all using a single Snapdragon processor, Visteon’s digital cluster, etc. In reality none of these HMI innovations triggered great interest like the digital speech assistant area did purely because of the aggression with which automakers are adopting those.

In conclusion the 2017 CES show did answer questions around commercial viability of some of the technologies and concepts demonstrated in 2016 and it added on the AI and deep learning heat. Frost & Sullivan is presently working on a very detailed report of CES 2017 that will cover all of these trends in greater detail by vehicle manufacturer, supplier, technology vendors and startups. The study will also feature several demonstration videos and interviews that Frost & Sullivan team undertook at the event. Frost & Sullivan’s Senior Partner and Head of Automotive Business Sarwant Singh and Partner and Head of EIA and Americas Automotive Practice Franck Leveque interview CEO of Renault Nissan Alliance Carlos Ghosn - https://www.youtube.com/watch?v=gYg2XNEugJ8. For any questions or queries please contact me at cpraveen@frost.com.

View Praveen Chandrasekar's blog

Comments (1)

By  Poornima Vijayan
Business Analyst Trainee, Cognizant

11 Jun 2017 01:36
1
  • page 1 of 1
Login or register to make a comment on this blog post

Help Desk

Full list of offices


For more information and general enquiries, contact Frost & Sullivan near you.

North America
tel: +1.877.463.7678

Select a location near you..