Researchers Bridge Gap Between Continuous Sensor Data and Automated Process Control

Business Process Management systems have traditionally handled discrete events in isolation, sidelining the continuous streams of data generated by sensors in cyber-physical environments. A new research effort tackles this gap by combining Signal Temporal Logic—a formal language for describing constraints across time-varying signals—with a real-time event processing architecture.

The Limitation of Traditional BPM#

Conventional business process models capture workflows as sequences of discrete steps: an order arrives, an invoice is generated, a shipment is dispatched. This abstraction works well for office workflows but breaks down in settings where physical systems are involved. A manufacturing line, a smart building, or an autonomous vehicle generates constant sensor readings alongside occasional milestone events. Existing process engines could monitor these hybrid scenarios only after the fact, checking whether the completed process conformed to rules. They could not actively respond to continuous conditions in real-time.

A Hybrid Execution Architecture#

The researchers propose a three-layer approach grounded in Complex Event Processing (CEP), a pattern-matching discipline for detecting meaningful patterns across streaming data. Their framework embeds Signal Temporal Logic predicates—formal statements about how signals should behave over time—directly into the execution layer. This enables the system to treat continuous sensor behavior as first-class participants in process control, not merely as monitoring sidecars.

In practice, this means the system can trigger activities based on continuous conditions ("start mixing when temperature exceeds 80°C") and enforce boundaries around process states ("reject any transaction if pressure remains outside safe range"). The declarative approach also keeps the specification human-readable and formally verifiable.

Why It Matters#

The bridge between hybrid specification and operational control has real implications for industries that straddle the digital and physical worlds. Smart manufacturing, IoT-enabled logistics, and robotics all depend on the tight integration of event-driven and signal-driven logic. By enabling systems to actively enforce constraints derived from both discrete and continuous behavior, this work moves hybrid process management from the audit lab into the control room.