Autonomous vehicle manufacturers depend on many expensive sensors to help their vehicles navigate road networks. How could Lunewave make sensor technology more accurate and cost-effective?
Combining its proprietary 3D-printed sensor technology with machine learning tools from IBM, Lunewave is developing an ultra-powerful sensor that will guide the autonomous vehicles of tomorrow.
Improvesthe accuracy of object detection in driverless vehicles, enhancing safety
Acceleratesinnovation by automating sensor design processes
Empowersautonomous vehicle manufacturers to reduce the cost of production
Business challenge story
Making dreams a reality
Driverless cars are no longer confined to the silver screen or pages of science fiction novels. By the early 2020s, many of the world’s leading automobile manufacturers and technology companies expect to release semi- or fully- autonomous vehicles. To achieve this ambitious goal, car manufacturers are working hard to develop driverless vehicles that can safely navigate their surroundings—both a technically complex and costly endeavor.
Finding a way to ensure that a self-driving car can transport passengers from A to B while adapting to ever-changing road conditions is one of the biggest challenges facing autonomous vehicle researchers. Even after innovators achieve this feat, the sensor technology used in autonomous vehicles is often so expensive that manufacturers will struggle to mass-produce an affordable driverless car.
Drawing upon decades of research into sensor technology, Lunewave developed a powerful and versatile sensor that holds the potential to reduce the cost of manufacturing autonomous vehicles.
Hao Xin, CTO of Lunewave and Professor of the Electrical and Computer Engineering Department at the University of Arizona, elaborates: “We have developed proprietary sensors based on the Luneburg lens design—a perfectly spherical lens that can detect light waves from all angles. Compared to other types of sensors, such as LiDAR and traditional Phased array antennas, a Luneburg lens has a number of distinct advantages: it can sense data across multiple bandwidths, detect objects in a 360-degree radius and communicate with multiple different devices simultaneously.
“Until now, manufacturing Luneburg lenses was a very time consuming and expensive process. But by harnessing the power of 3D-printing, we can make highly accurate Luneburg lenses in minutes or hours and at a fraction of the cost of traditional manufacturing processes.”
Recognizing the enormous potential of its sensors, Lunewave set out to refine its technology for the emerging autonomous vehicle industry.
“One Lunewave sensor is powerful enough to replace up to 12 of the sensors that are currently being used in autonomous vehicles,” explains Professor Xin. “But to make our sensor market-ready, we first needed to train it to avoid interference from other sensors on the road and, most importantly, learn how to classify different types of objects, such as the difference between a pedestrian, cyclist and a road sign.”
Developing a feel for the road
Seeking a platform to support its product development goals, Lunewave teamed up with IBM, who recommended using IBM Watson® Studio (formerly IBM Data Science Experience) to advance the company’s proprietary sensor technology.
Professor Xin continues: “We decided to work with IBM as we wanted to draw upon the expertise and advice of their data science experts. We are currently working with IBM engineers to calibrate our sensors using the machine learning capabilities of the IBM solutions and our own algorithms. Our goal is to train our sensors to screen out interference from other radar devices, learn how to classify different types of objects and understand how to avoid typical danger scenarios that drivers encounter on the road.
“With Watson Studio, we can harness advanced data science and analytics tools to achieve real-time object detection—which will enable the sensor to spot and respond to objects as they move around a driverless vehicle. This will make the cars themselves much safer on the road. The fact that these solutions could support this vital capability made deciding to work with IBM a no-brainer.”
As well as using IBM solutions to detect objects more accurately, IBM Watson™ Studio enables engineers at Lunewave to automate and streamline the design of new types of sensors.
“The combined power of 3D printing and machine learning technology is tremendous,” says Professor Xin. “For instance, 3D printing offers engineers millions of ways to bring together new combinations of materials in intricate designs. However, with such an increase in the number of potential designs, choosing the most effective blueprint and material combination for a new product becomes much harder and more time consuming.
“To optimize and streamline the design of new sensors, we can utilize the machine learning and AI features of Watson Studio to search through millions of potential blueprints automatically and find the optimal construction. While we haven’t made use of this automated design process with our Luneburg lenses, we're really excited to use Watson Studio to further refine our other models in the months ahead.”
Innovating in the fast lane
With IBM Watson Studio helping to drive the development of its cutting-edge sensor technology, Lunewave can focus on accelerating innovation in the autonomous vehicle sector.
Professor Xin observes: “The feature-rich Watson Studio platform equips us with all the tools we need to train and develop our sensors to screen out radar interference and navigate road networks safely and efficiently in real time. Thanks to our IBM solutions, each day we move a little closer to refining our sensors, so that they become more effective at spotting and avoiding hazards on the road than a human driver ever could be.
“What’s more, because Watson Studio is such a flexible platform, we can apply them to help us solve many different problems. As a result, we are able to streamline the sensor design process and formulate original, cutting-edge blueprints for new products—helping us stay at the tip of the spear of object detection technology.”
At present, Lunewave is focusing on developing its Luneburg lenses for the autonomous car market; however, the technology could be applied to a much wider range of use cases.
Professor Xin concludes: “By advancing sensor technology with the support of IBM experts and IBM technology, we are laying the foundation for automotive manufacturers to produce autonomous vehicles equipped with reliable, robust sensor technology at an affordable price.
“Our first stop is to revolutionize the autonomous vehicle market, but after this junction we intend to develop sensors that can be used in other applications such as drones and other types of Unmanned Aerial Vehicles. We truly believe that the future of travel lies in flying cars, and IBM technology will help us develop the sensor technology to make this a reality.”
Founded in 2017 with ties to the University of Arizona, Lunewave is a technology company developing disruptive antenna and radar sensor technology by leading experts in millimeter wave frequency engineering. Products are geared toward a variety of markets including automotive, telecommunications, aerospace, and research for applications such as autonomous transportation, wireless communications, and drones. For more information, visit lunewave.com.
Take the next step
IBM Analytics offers one of the world's deepest and broadest analytics platform, domain and industry solutions that deliver new value to businesses, governments and individuals. For more information about how IBM Analytics helps to transform industries and professions with data, visit ibm.com/analytics. Follow us on Twitter at @IBMAnalytics, on our blog at ibmbigdatahub.com and join the conversation #IBMAnalytics.