IBM said that it has patented a machine learning technology that defines how to shift control of an autonomous vehicle between a human driver and a vehicle control processor in the event of a potential emergency.
+More on Network World: IBM on the state of network security: Abysmal
Basically the patented IBM system employs onboard sensors and artificial intelligence to determine potential safety concerns and control whether self-driving vehicles are operated autonomously or by surrendering control to a human driver.
The idea is that if a self-driving vehicle experiences an operational glitch like a faulty braking system, a burned-out headlight, poor visibility, bad road conditions, it could decide whether the on-board self-driving vehicle control processor or a human driver is in a better position to handle that anomaly. If the comparison determines that the vehicle control processor is better able to handle the anomaly, the vehicle is placed in autonomous mode,” IBM stated.
“The technology would be a smart wingman for both the human and the self-driving vehicle,” said James Kozloski, manager, Computational Neuroscience and Multiscale Brain Modeling, IBM Research and co-inventor on the patent.
+More about IBM research on Network World: IBM: Next 5 years AI, IoT and nanotech will literally change the way we see the world+
Kozloski also noted another recently issued IBM patent that will help automated cars and human-operated vehicles more safely interact in the future.
In a nutshell, this technology uses what’s called automatic driver modeling which to watch for the behavior of a person driving a car, such as use of turn-signal, speed and other behaviors to make assumptions about that car and pass those observation onto other autonomous cars in the vicinity.
A model of the driver is generated based on the behavior patterns of the driver measured by the one or more sensors of the autonomous vehicle. Previously stored behavior patterns of the driver are then retrieved from a database to augment the model of the driver. The model of the driver is then transmitted from the autonomous vehicle to nearby vehicles with autonomous interfaces, Kozloski said.
“The whole idea with both patents is to enable a safer environment for humans,” Kozloski said.
IBM says it has patented numerous inventions that, among other things, can help vehicles become:
- Self-learning – powered by cognitive capability that continuously learns and gives advice based on behavior of the driver, passengers, and other vehicles
- Self-socializing –connecting with other vehicles and the world around them
- Self-driving –moving from limited automation to becoming fully autonomous
- Self-configuring– adapting to a driver’s personal preferences
- Self-integrating –integrating into the IoT, connecting traffic, weather, and mobility events with changing location
Indeed, IBM is deeply involved in the self-driving car arena. In December, the company said it would collaborate with BMW and its own cognitive computer whiz Watson to develop self-driving cars that can adapt to driver preferences.
The IBM study, “Automotive 2025: Industry without borders,” amassed interviews with 175 executives from automotive OEMs, suppliers, and other leaders in 21 countries and found that by 2025 cars will be able to learn, heal, drive and socialize with other vehicles and their surrounding environment.
Some of the study’s more interesting observations included:
- 2025, the vehicle will be sophisticated enough to configure itself to a driver and other occupants.
- Fifty-seven percent believe vehicle “social networks” would be in place where vehicles would communicate with each other, allowing vehicles to share not only traffic or weather conditions, but information specific to a given automaker. For instance, if a vehicle was experiencing some type of problem not recognized before, it could communicate with other vehicles of the same brand to seek help on what the issue might be.
- Analytics capabilities will help vehicles identify and locate issues, schedule fixes and even help other vehicles with similar problems with minimal impact to the driver.
- Like other smart devices, the vehicle will be an integrated component in the Internet of Things (IoT). It will collect and use information from others concerning traffic, mobility, weather and other events associated with moving around: details about driving conditions, as well as sensor-based and location-based information for ancillary industries, such as insurance and retail.
- Seventy-four percent of respondents said that vehicles will have cognitive capabilities to learn the behaviors of the driver and occupants, the vehicle itself and the surrounding environment to continually optimize and advise. As the vehicle learns more about the driver and occupants, it will be able to expand its advice to other mobility services options. The report also underscores considerable skepticism about fully autonomous vehicles—where no driver is required and the vehicle is integrated into normal driving conditions. A mere 8% of executives see it becoming commonplace by 2025. Moreover, only 19% believe that a fully automated environment—meaning the driving system handles all situations without monitoring, and the driver can perform non-driving tasks—will be routine by 2025.
- Eighty-seven percent of the participants felt partially automated driving, such as an expansion of today’s self-parking or lane change assist technologies would be commonplace. Moreover, 55% said highly automated driving, where the system recognizes its limitations and calls driver to take control, if needed, allowing the driver to perform some non-driving tasks in the meantime, would also be adapted by 2025.
Check out these other hot stories: