Inside Unmanned Systems provides actionable business intelligence to decision-makers and influencers operating within the global UAS community. Features include analysis of key technologies, policy/regulatory developments and new product design.
Issue link: http://insideunmanned.epubxp.com/i/763107
43 unmanned systems inside December 2016/January 2017 prediction. Knowing the risk right now is good but it's better to be able to predict it so we can actually do something against a hazardous situation. If we have an overall safety require- ment, for example, we want to do better than 1 fatality in 100 million miles driven. If we have a metric to analyze and quantify the integrity of each one of these components (that make up the driverless vehicle), we can allocate the overall safety requirements to each one of these components and we can design each one of these components for safety. That design for safety will accelerate the development of driv- erless cars." The Industry's Perspective For more than 20 years NovAtel has been the leader in high precision GNSS technology. Re- cently, the company formed a Safety Critical Systems group to create positioning products that meet the performance and safety require- ments of autonomous vehicles for a variety of industries—and that includes driverless pas- senger cars. While GNSS will serve as the absolute po- sition, velocity and time generator for these driverless vehicles, Auld said, it won't be the only positioning source. He agrees driverless cars will need a fusion solution of multiple sensors. GNSS will need to be available at all times, no matter the weather conditions, and able to support the necessary position- ing accuracy. Driverless vehicles will move to multi-frequency and multi-constellation sensors, with PPP (Precise Point Positioning) available through PPP correction services, like TerraStar, that can deliver corrections over satellite and over the Internet. One of the challenges, though, is it often takes PPP tens of minutes to get to the 5 to 10 cm level. A number of companies are working to re- solve this, so that limitation should be large- ly removed in the next few years. "These concepts of integrity, continuity, accuracy and availability have to be bal- anced and understood in an automotive use case," Auld said. "Dealing with threat models like multipath are completely different and much more challenging from an automotive perspective than from an aviation perspec- tive. These concepts will have to be extended into the ground applications such that on the most basic level the receiver cannot provide a misleading set of information. If the receiver is providing a good solution, it needs to be trustworthy enough to say 'Yes, here's your answer and it is good and you can trust me to a given integrity risk level.' If it is not able to provide that solution, it needs to be able to provide that information as well. What you don't want to have is the receiver's output providing an indication it is good and trust- worthy when it is not." Moving Forward Driverless cars need accurate positioning that is also safe—and that isn't a simple problem to tackle. Researchers and manufacturers will continue to look for the best solutions to move the autonomous driving industry toward mak- ing driverless cars a reality on our roadways. To get there, products need to be built with safety requirements in mind, and while the car industry faces its own unique challenges, a lot can be learned from what's already been done in the aviation industry. "This isn't just a matter of designing a new receiver and taking it to the test lab," Auld said. "We have to design it from the ground up for the receiver, the antenna and the net- work to hit the high safety standards for the autonomous industry and any of the other in- dustries. The additional challenge of course is the automotive world has significantly higher volumes and significantly lower cost expecta- tions. High precision is not currently available in that cost structure so there's a lot of work to do to bring those solutions in at the cost and volume levels the automotive industry is expecting." HOW TO WATCH THE WEBINAR Visit http:// www.insidegnss.com/ webinars and register