Preparing for the Mobility Revolution
Connected Services Are Revolutionizing the Mobility Experience
The mobility revolution we are currently experiencing, the most disruptive since the automobile was introduced about 130 years ago, is the result of four mega-trends: electrification, autonomy, shared mobility and connectivity. Vehicles are becoming increasingly connected to their users, owners and pedestrians, to each other, to the infrastructure and more. A myriad of value propositions have already emerged and many more should be expected. They benefit safety, user experience or productivity. Let’s see how this is unfolding for various categories of value propositions.
Enabling the Safe Deployment of Highly Automated Vehicles
Whereas connectivity can be a “nice to have” for certain use cases, it will be necessary in various semblances to maximize the safety of highly automated vehicles. Vehicle to vehicle (V2V) communication can address blind spots when cameras, lidar and radars cannot see around the corner, and provide overall redundancy. Savari, Veniam or Autotalks have developed different parts of the solution to this problem. They not only provide the location of other vehicles, but also potentially information as to their behavior, e.g., acceleration or braking.
Vehicle to infrastructure (V2I) communication enables a dialog with traffic management systems and road markers. This will eventually allow for traffic to flow more smoothly thus more efficiently, for emergency vehicles to be given priority over regular traffic, and potentially to help autonomous vehicles (AVs) precisely locate themselves. A few corridors, such as El Camino Real in Silicon Valley, have been fitted with DSRC-based equipment to test the technology. BMW is also deploying a solution developed by Connected Signals which provides current and predictive information on signal status (green/yellow/red). This allows drivers to adapt their speed and drive more smoothly.
At least for the foreseeable future, AVs will require 3D HD maps to precisely locate and guide themselves. Such maps are likely to be crowdsourced and updated in realtime in order to reflect changes, such as road construction. Gaps observed vs the onboard reference map will be pushed to the cloud, confirmed with similar data from other vehicles and pushed back to all vehicles in the concerned area. For instance, Mobileye is collaborating with BMW, VW and Nissan on such a solution. By end 2018, they expect to have 2 million vehicles capable of crowdsourcing map data.
Drivers and pedestrians are accustomed to informal communication. In an AV world, pedestrians must be made aware of vehicles’ intentions. Conversely, vehicles must identify all pedestrians who may get in their paths, whether they are visible or not. Autonomic, which was acquired by Ford earlier this year, is incorporating pedestrians into their cellular V2X platform.
Driver and Vehicle Monitoring Enhance Safety and Cost
Vehicles already generate a large amount data due to a multitude of sensors. Intel predicted that an AV will generate 4 TB of data each day. Currently, this data is by and large processed onboard for the purpose of optimizing vehicular operations. However, a number of companies are already taping the data available on the vehicle’s CAN bus (via aftermarket dongles installed on the OBD port) or adding extra sensors, e.g., cameras. The data is exported via 4G and processed remotely or sold to third parties. This opens the door to a myriad of services.
Some of these services relate to safety, such as Zendrive or i4drive, which analyze driving behavior and provide feedback to both drivers and fleet managers. On top of this, Nauto alerts drivers when their behavior puts them at risk in dangerous situations, e.g., vehicle breaking ahead. Safe driving habits can mean lower premiums via Usage Based Insurance, which is the main focus of Octo Telematics.
Vehicle diagnostics and predictive maintenance is critical, not just to reduce operating costs, but also to remotely diagnose vehicles when the user is not the owner. In addition, it can provide leads to service networks. Companies like Mojio or Voyomotive address such opportunities. Carfit (which I am advising) uses vibration AI to diagnose systems that are not fitted with sensors with the same objective. The monitoring of a vehicle’s parameters creates a richer history than that built on dealer visit records (e.g. Carfax), which will enable fact-based fleet valuation. VINchain plans to build a blockchain-based history of individual vehicles to do just that.
Tesla has been collecting vehicle output since the introduction of Model S. The resulting data lake provides a huge resource that leads to engineering improvement as well as business performance. The rest of the industry is following suit, as natively connected vehicles will represent about 80% of global sales vs 40% today. Yet this does not suffice. Corporate culture and expertise at incumbent players must evolve to leverage the full potential of this massive data.
A Plethora of Extra Convenience and Services
If safety remains the most critical objective, customer benefits allow for differentiation, thus extra value. By introducing over-the-air (OTA) updates in 2012, Tesla was not only able to fine-tune its software overtime — without the need for customers to come to the dealer — but also to add new features on the fly. Examples include Autopilot updates, the recent reduction in braking distance on Model 3, the temporarily range increase in an emergency situation or the addition of voice command. Movimento now offers such OTA services to all OEMs.
Connected vehicles also offer extra convenience. While we still own cars, remote access enables services such as dropping your Amazon parcel in your trunk, servicing, refueling or cleaning your car while you work, etc. SmartCar is working with BMW, Hyundai and PSA on such a solution. The understanding of human behavior inside the cabin will provide insights first on the driver (awareness, ability to take over the driving role) and then on passengers. Identifying their sex, gender, emotion, position or activities inside the cabin will enable a series of features including targeted promotion. Eyeris (which I am advising) has a production-ready solution for both drivers and passengers.
With these services comes the need for secure payment. Vehicle-specific platforms are coming to life to allow for the payment of tolls, parking, EV charging, foods or other in-car services without human intervention. Several players are developing solutions, including Tier 1 supplier ZF and IBM who are jointly developing ”Car eWallet.” GM is rolling out Marketplace, which enables access to third party service providers directly on the vehicle’s head unit, as well as payment. Xevo, which powers Marketplace, just announced a similar deal with Hyundai to develop their Digital Wallet.
Drivers and passengers alike expect continuity between their smartphones, vehicle displays and home/office. To this end, Apple and Google are already widely present in the cabin respectively with CarPlay and Android Auto. Whereas Apple has so far allowed very few proprietary apps, the giant will open up and include Google Maps and Waze with iOS12 later this year. CarPlay and Android Auto come with their digital assistants, i.e., Siri and Google Assistant. However, Amazon’s Alexa seems to have taken the lead when it comes to in-car assistant. A few startups also work on Natural Language Processing (NLP) for cars, such a Xbrain, German Autolabs or Mobvoi.
Privacy and Monetization
Who owns the data? The user, the OEM, the solution supplier or any 3rd party? How is data transfer controlled? Europeans are more protective of their personal data than Americans, who mainly put free services ahead of privacy. GDPR, introduced last May in the EU, puts citizens at the center. They own and have full control over the use of their personal data — a user’s consent is required before such data can be shared. Since the USA lacks the equivalent of GDPR, early signals seem to indicate that the GDPR may become the de-facto global rule for any company that does business in Europe.
Once data ownership is sorted out, business can be made from monetizing data. A majority of the companies mentioned above make money off of their data in some ways. However, marketplaces have been created for the sole purpose of trading automotive data, connecting sellers and buyers. Companies behind them includes Caruso in Europe or Otonomo in the USA. Such companies often process or augment the data through analytics in order to increase their value.
How about the Tech behind these Services?
Whereas WiFi and DSRC are likely technologies for V2V and V2I, the latter has only been deployed so far by Toyota and Cadillac. This is a major hinderance as it takes two to communicate. The large amount of data sent and received by vehicles will use 4G, then 5G when deployed — Audi has already announced they will roll out 5G-connected vehicles starting in 2020. In specific use cases where a narrow bandwidth is sufficient and energy is scarce (e.g. vehicle tracking), the ultra narrow band modulation tech deployed by Sigfox or LoRa will make sense. As for, vehicle to pedestrian communication, cellular is main contender though LiFi, an LED-based tech deployed by Zero.1, may be an alternative. Whichever the communication, a robust cybersecurity layer is absolutely essential, but that’s a whole other topi, which is not addressed here.
Lastly, large amounts of data will means large bandwidth and high cost. For this reason, I expect that processing data at the edge will become more prevalent, unless its monetization potential far exceeds the cost of beaming it. However, a side benefit from processing more data locally will be that more personal data will remain inside the vehicle, e.g. the dialog you have with your digital assistant will not be shared!
Managing Director, Orsay Consulting