GM Gets Personal With IBM Watson: Cars get OnStar AI System

General Motors Corporation is enlisting the aid of IBM’s Watson AI platform to create OnStar Go, a smart interactive assistant for automobiles. While the basic GM OnStar platform has been around for 20 years, every car maker today is working on enhancing its network-connected entertainment and vehicle management system. This is a good opportunity for both companies. By the end of 2016 GM expects to have 12 million OnStar connected vehicles on the road worldwide. OnStar’s primary mission is to supply safety help and communications by connecting to a cellular network, and this role is expanding as the market continues to develop.

OnStar Go will appear early next year in more than 2 million GM vehicles using 4G cellular service. It is a subscription service providing entertainment and safety capabilities. It uses Watson’s cognitive capabilities in a manner similar to a home smart personal assistant such as Amazon Echo or Google Home. OnStar Go bills itself as the first “cognitive mobility platform”.

The capabilities that IBM Watson brings to this platform will include a range of similar services to those provided in varying degrees by smart phones and smart home assistants, plus others that involved its capacity to monitor and control the mechanisms of the vehicle. OnStar Go will be able to help drivers with a growing range of services that demand knowledge of location, personal habits, and vehicular control.

Start with the Car

According to IBM’s press release:

“Combining OnStar’s industry leading vehicle connectivity and data capabilities with IBM Watson APIs will create experiences that allow drivers and passengers to achieve greater levels of efficiency and safety. These experiences could include avoiding traffic when you’re low on fuel, then activating a fuel pump and paying from the dash; ordering a cup of coffee on the go; or getting news and in-vehicle entertainment tailored to your personality and location in real time.”

Cognitive capabilities bring a range of new possibilities for personalizing the driving experience. Initial offerings seem fairly similar to what is already available; but the capability to learn from experience and store vast amounts of data makes it possible to create a more highly customized experience as well as presenting an opportunity to extend marketing and branding possibilities.

“On average, people in the U.S. spend more than 46 minutes per day in their car and are looking for ways to optimize their time,” said Phil Abram, Executive Director, GM Connected Products and Strategy. “By leveraging OnStar’s connectivity and combining it with the power of Watson, we’re looking to provide safer, simpler and better solutions to make our customers’ mobility experience more valuable and productive.”

As a subscription product, OnStar Go will provide GM with a continuous revenue stream in the entertainment and services industry. The platform also provides plenty of opportunities for learning how people prefer to set up their environment and where they would like to go. This data is useful in creating better personal assistants. IBM does not as yet have a personal assistant tool competing in the Google/Apple/Amazon space; it is clear that such platforms are the way of the future. Voice-operated request systems with cognitive interactions that decode naïve requests and formulate a complex response based upon database and Internet search capability will shape home and office environments for years to come.

Caveat Emptor

There are items of particular note in providing cognitive links to automobile functions. The OnStar system has potential to stop the car or manipulate its controls. It provides parking capabilities and navigation capabilities which will grow in importance. This kind of control presents security risks which have already been demonstrated last year with a pre-Watson version of OnStar compromised to remote control a Chevrolet Impala, as reported by the television program 60 Minutes last year. While security problems get patched, the possibility of sophisticated access to vehicle systems will always remain. Watson will require large amounts of data, some of it personal; this data can likely be hacked to provide information about whereabouts and intentions of the user; at the same time it should be able to provide more sophisticated intrusion response. Another issue is mobility. A continuous attachment to the cellular network as a requirement for operation could create possibility for security compromise or interruption of complex actions when the network fails. With the connection to vehicle systems, this could create dangerous vulnerabilities that might impact the driver. Additionally, the data collected will likely be provided to third parties for marketing, making it possible for outsiders to obtain special knowledge for marketing–most of which will be that benign, but some of which may be pernicious in a moving vehicle.

On to the Future

The ability to provide smart assistance within an automobile is probably the most difficult of the tasks to be expected of a personal assistant. These systems have been of marginal value in the home with relatively little to control. Automobiles are different. They are a transportation system in motion in a sophisticated environment, with innumerable demands–some of which might include personal comfort, entertainment, navigation, purchase of outside services, fueling, environmental issues, lighting, safety, and response to the unexpected.

This is a whole new breeding ground for smart assistance. While all major automobile manufacturers are working upon similar types of help, particularly as the era of autonomous vehicles approaches, the Watson approach brings in some new possibilities. The complex, cognitive learning system with integrated database capabilities and search makes it possible to craft a more interactive user interface as well as operate more deeply within the user environment. This will create a challenge for other players in the smart automobile as well as in the personal assistant sectors.

The personal assistant industry is developing quickly as visions of the smart home come together. Companies understand that they need to be visionary here. Those who establish a viable platform and are able to demonstrate success in this category will be able to apply it across the board to the consumer and commercial environments. This Holy Grail of marketing and interactive assistance will be fundamental to the complex operation of the Internet of Things; it will bring on the age of personal assistants, robots, and the advent of ubiquitous intelligence in processes.


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.