Dynamically optimizing experiment schedules of a laboratory robot system with simulated annealing

Area of Science:

• Robotics and Automation
• Biochemistry and Structural Biology
• Computational Science

Background:

• Integrated laboratory robot systems require efficient scheduling for continuous operation.
• Protein crystallization experiments necessitate timely imaging at multiple time points.
• Dynamic rescheduling is crucial for accommodating new experimental plates in real-time.

Purpose of the Study:

• To develop and optimize a scheduler for an always-on integrated laboratory robot system.
• To enable dynamic rescheduling of imaging tasks for protein crystallization experiments.
• To balance schedule quality with computational speed for real-time adjustments.

Main Methods:

• Developed a scheduler based on a simulated annealing algorithm.
• Integrated a linear programming solver into the objective function for optimization.
• Conducted extensive computational simulations to test various algorithm configurations.

Main Results:

• Identified an optimal configuration for the simulated annealing algorithm.
• Achieved scheduling results in under 60 seconds, meeting real-time requirements.
• Demonstrated the scheduler's ability to dynamically reschedule imaging tasks.

Conclusions:

• The developed scheduler effectively manages dynamic imaging tasks in an always-on laboratory robot system.
• Simulated annealing with linear programming provides a computationally efficient solution for complex scheduling problems.
• This approach enhances the flexibility and efficiency of protein crystallization studies.