John P. Lehoczky

John P. Lehoczky
John P. Lehoczky

Biography

The groundbreaking work of John P. Lehoczky, Ragunathan Rajkumar, and Lui Sha on developing the generalized rate-monotonic scheduling (GRMS) theory has revolutionized the modern practice of real-time system design by ensuring that critical tasks can be guaranteed across a wide range of real-world applications. Building upon the RMS theory introduced by Liu and Layland in 1973, the trio developed and refined GRMS theory over two decades to provide predictability, efficiency, and flexibility in scheduling complex concurrent real-time tasks that regular RMS could not satisfy. GRMS transformed the development of real-time systems from what was traditionally a hand-crafted, error-prone process into a scientific engineering discipline for building real-time systems that are pervasive in applications including aerospace, defense systems, transportation, process control, manufacturing, and medical systems. They constructed the first exact test to determine whether periodic tasks could be scheduled to meet general real-time constraints; expanded the scope of this test to include aperiodic tasks; developed robust protocols to synchronize real-time tasks; and determined the priority levels required to guarantee real-time behavior. GRMS made real-time design possible for complex, industrial-scale systems that exhibited rich behaviors while meeting essential, safety-critical design constraints. Their work made it possible to compute both the average-case and worst-case behaviors with useful mathematical precision. Not only does GRMS contribute to keeping development costs in check, leading to billions of dollars in savings and increased efficiencies being accrued, it also helps assure that expensive mission failures do not occur. GRMS has been incorporated into IEEE real-time software standards such as POSIX Real Time Extension and hardware standards such as IEEE Futurebus+. GRMS has been implemented in major projects including a wide variety of U.S. Department of Defense systems; the International Space Station; global positioning systems; the Mars Pathfinder; and, more recently, in autonomous vehicles.

An IEEE member, John P. Lehoczky is the Thomas Lord University Professor of Statistics and Mathematical Sciences, Carnegie Mellon University, Pittsburgh, PA, USA.