SARAS: Autonomous robotic arms with a cognitive system to assist surgeons in the operating room

As part of SARAS, next-generation surgical robotic systems have been created so that a single surgeon can execute robotic minimally invasive surgery (R-MIS) without the need of an expert assistant surgeon. This increases the efficiency of the social and economic resources of a hospital and guarantees the same safety level for patients.

The concept of robot that has been worked on consists of an integral solo-surgery system, through the integration of two additional assistive robotic arms (which assume the tasks of assistant surgeon) that are autonomous, and a cognitive and cooperative artificial intelligence core with a surgical robotic system. To achieve this, the auxiliary robots have to learn the basic surgical procedure. They must be able to determine which phase of the process is being carried out, so that they can decide which assistance action must be undertaken once the action of the main surgeon has been completed. This assistance could consist in moving an organ or holding a suture needle.

The concept required the development of specific technologies for its implementation:

• The design of a perception module, which will infer the state of the procedure and the actions carried out by the surgeon who tele-operates the robotic arms or uses standard laparoscopic instruments. This module is also responsible for understanding the complexity of the surgical area by reconstructing, labelling and tracking all of its elements as observed by the available sensors.
• The design of a cognitive control model that predicts the following steps in the procedure based on data on real interventions in the structure of complex laparoscopic procedures to identify anomalies, understand the surgeon’s actions (awareness of the situation) and their future needs (decision making).
• The design of a low-level control architecture for multiple robots, in which high-level orders are outlined in orders for the robotic arms. This required the design of an advanced planning and navigation module that could translate the autonomous decisions made by the cognitive module into the appropriate trajectories for the laparoscopic tools mounted on SARAS assistive robotic arms.
• The development of the two assistive robotic arms, instead of the assistant surgeon, with a capacity to guide the standard and specific laparoscopic instruments for the clinical procedure.