Smart, Privacy-Enhanced Edge Computing
From autonomous cars to strategic defense systems, emerging smart technologies must achieve near-instantaneous response times to operate successfully in dynamic, fast-paced environments.
Kenneth P. Birman, Computer Science, is creating a new cloud computing infrastructure called Cascade, which will shift machine intelligence to a network’s edge—that is, toward small compute units, equipped with sensors and actuators, that can react in real time to rapidly evolving situations.
Cascade will host applications much as cloud computing does today. Essentially a distributed, high-assurance, key-value storage infrastructure with speeds thousands of times faster than a standard file system, Cascade will support edge computing with machine-intelligence capabilities—creating a new option for smart solutions in domains such as agriculture, energy, transportation, and national defense. Cascade might support a smart highway, for example, that gathers data through sensors on cars. The system could respond to changing road conditions by alerting drivers to unexpected hazards or by designating express lanes for safe travel at high speeds.
Cascade runs on Cornell University’s record-setting open-source Derecho software platform and couples remarkable performance with unusually strong guarantees. This research will enhance Cascade with privacy-preserving architecture for applications that involve sensitive or secret data. The system will maintain sensitive data close to its point of capture and will eliminate such data as soon as it is no longer needed. Privacy enhancements will guarantee that only non-sensitive information leaves the Cascade-provided computing environment.