Online real-time platform for microrobot steering in a multi-bifurcation

Abstract

The study of methods to control a collection of ferrous microparticles (microswarm) within an electromagnetic field is a branch of medical robotics that has many promising applications. To this date, most simulated data is obtained from accurate but time-consuming simulations of which the biggest disadvantage is the removal of the human from the control loop. In practice a human in the loop will in many cases be controlling the microswarms. In order to re-introduce the human aspect into the control loop, this paper describes the development and validation of a real time simulator. By satisfying the real time requirement, we can record users interactions with the microswarm in a similar way that the user would react with real world particles. The context of each simulation is for the user to direct a microswarm through a multi-bifurcation towards a selected outlet. The percentage of particles reaching the selected outlet is considered as the success metric. The model has been verified against previous experimental data.The platform showed 8% deviation from the experimental data. Parametric studies were then undertaken, providing an initial data set to analyse different real time articulation strategies. It was found that the fluid flow velocity made the most difference to the success of the user. With lower fluid flow velocities of 0.001m/s to 0.005m/s, close to 100% of particles reached the chosen outlet. This drops to around only 30% of particles reaching the chosen outlet at a fluid flow velocity of 0.25m/s. The other parameters particle size and capped magnetic force produced a much lesser variation in results, with the maximum and minimum recorded numbers of particles reaching the chosen outlet for both tests only ranging 20% each. The platform, being generic, would also be appropriate as a training tool.