| Hearken to this text |
Researchers at a college have devised a plan to enhance robot reliability by developing flexibility in coping with diverse levels of vulnerability. | Supply: Adobe Inventory
Scientists from the University of Rochester, Georgia Institute of Technology, and the Shenzhen Institute of Synthetic Intelligence and Robotics for Society have unveiled a pioneering approach to safeguarding robotics against vulnerabilities at a fraction of the cost.
Tens of thousands of people are expected to be without a home by 2025, while autonomous vehicles are currently generating billions of dollars in annual revenue. As such, ensuring product quality, consistency, and safety become paramount concerns for consumers, manufacturers, and regulatory authorities alike.
Despite this, programs designed to safeguard autonomous machines’ hardware and software from malfunctions, attacks, and other failures also increase costs. The prices are determined by a combination of factors, including efficiency options, power consumption, weight, and the utilization of semiconductor chips.
Researchers caution that the current trade-off between security overhead and defenses against vulnerabilities stems from a rigid “one-size-fits-all” approach to safety. Here is the rewritten text:
The authors’ study revealed a novel approach that effectively addresses a wide spectrum of vulnerabilities within autonomous systems, ultimately enhancing their reliability and reducing management costs.
Yuhao Zhu, a respected affiliate professor at the university, highlighted one notable example: Tesla’s innovative approach to equipping its vehicles with two Full Self-Driving (FSD) chips in each car. While this redundant system provides added security by having a backup in place should the primary chip fail, its implementation more than doubles the cost of the automotive chip, making it a significant trade-off.
While discussing security measures, Zhu noted that he and his university students implemented an all-encompassing approach to protect against both hardware and software vulnerabilities, thereby ensuring effective allocation of security resources.
The researchers’ breakthrough innovation revolves around crafting a tailor-made approach to safeguarding against the rise of automation?
A comprehensive framework of diverse software programs and hardware-based safeguards for robust autonomous systems? Will you have a 5G-enabled smartphone by the end of 2023? The answer depends on several factors, including your carrier, phone model, and geographic location.
“The underlying concept is to implement distinct safety protocols for various components of the system,” Zhu explained. The refinement of a strategy should ultimately hinge on the intrinsic characteristics of both the software and hardware. Developing distinct safety protocols for both the front-end and back-end of the software stack is essential.
The initial mention describes the autonomous car’s software program stack focusing its front end on sensing the environment through components such as cameras and lidars, while the back end processes this data, plans a route, and sends instructions to actuators.
.
