Presentation
Runways and Robots: An Analogy Awakened
DescriptionRobotic surgery has recently seen a rapid rise at a global level. Several robot types are now in clinical practice, with da Vinci, Versius and Hugo robots in the UK. Previously, robotic training was acquired by peri-consultant surgeons. Today, training will need to start earlier in a surgical career. This creates the need for a robotic curriculum to train surgeons of varying experience and gaining access to several robot types. The analogy with aviation is fruitful here. Aircraft are classified into ‘types’. Significant differences between aircraft requires specific training, culminating in a ‘type rating’. As a robotic surgeon and pilot, I apply my experience to draw on this analogy to identify learning that may be transferable across fields and so derive principles that would be useful in curriculum design. We cover the human-machine interface, team training, simulation, control and automation.
Cockpit and console
Pilot training starts with gaining the private pilot’s licence (PPL). The type rating for light aircraft is a broad category and includes a range of aircraft with largely similar cockpit layout and function. As aircraft become larger and more complex, the type rating contains fewer aircraft, such as a single aircraft (e.g. A380), or two aircraft designed with cockpit commonality to form a common type rating (e.g. Boeing 777 and 787). Similarly, in surgery, early stage training is in open and laparoscopic surgery. With this prior experience, robotic training has traditionally focused on using a particular robot type to perform an operation known to the learner. With less prior experience, there is a dual focus on both learning to use the robot as well as to perform the operation. With prior robotic experience, the concept of ‘transfer’ applies. Prior surgical and robotic experience therefore modulate the needs of a robotic curriculum into the following components: core (cross-platform, cross-specialty), specialty-specific, platform-specific, and platform-transitional.
Transfer
The concept of ‘transfer’ from the aviation industry helps explain this. Moving from one aircraft type to another can affect cockpit workload and performance through negative transfer or habit interference, conceptualised in a ‘transfer matrix’. The real-world significance of ‘transfer’ is highlighted by the several studies reporting a concurrent use of or serial transition from da Vinci. The impact of prior console expertise on performance has largely shown positive transfer. This is considered further below.
The team
Crew resource management (CRM) is formal mandatory team training used in aviation since the 1980s, focusing on non-technical skills. Non-technical skills training has also been delivered in surgery as the Non-Technical Skills in Surgery (NOTSS) course. Future iterations of this kind of course may bring a core (cross-platform) robotic element. Of note, the operating surgeon sits remote from the bedside, somewhat analogous to the radio-communication in aviation. Aviation communication tools such as standardised phraseology and read back may be particularly useful.
Human-machine interface
Dual control is standard in many aircraft; in general aviation, for training, and in commercial aircraft, to facilitate alternation of role. In robotic surgery, da Vinci offers a dual console option. Evidence is mixed regarding the benefit of dual over single console, noting loss of ‘embodied guidance’ with dual console. Other aspects of console and control layout can affect both platform-specific and platform-transitional components. Regarding console controls, da Vinci and Hugo have hand and foot pedal controllers, whilst Versius has only hand controllers. This increases hand activity in Versius users, but potentially reduces error in laterality of energy application because the energy control input is located on the same hand-controller that operates the instrument. This is an important difference, particularly relevant for a Versius-to-pedal-platform transition.
Another more general consideration is the lack of haptic feedback or force-feedback. In aviation, light aircraft have a direct mechanical coupling between cockpit controls and aerodynamic control surfaces. Force and vibro-tactile feedback in the cockpit is therefore an implicit property of the system. Modern airliners have flight-by-wire (FBW), which completely replaces the physical connection with electro-mechanical coupling. This is supplemented with flight envelope protection, which reduces pilot control inputs such that the aircraft remains within allowable limits. However, this envelope protection is not perceived by the pilot. This lack of feedback may have contributed to aviation accidents and has led to research in conveying flight envelope protection through haptic feedback in active control devices. By analogy, similar changes have occurred in surgery. Haptic feedback naturally occurs in open surgery, in direct proportion, and to a lesser extent in laparoscopic surgery, where forces are multiplied by the lever arm. Robot surgery is in effect surgery-by-wire (SBW). Robot surgery currently provides no haptic feedback, nor is there envelope protection. Though there is currently no clearly defined ‘surgical envelope’, dimensions of the envelope could include traction force and spatial reach, to constitute envelope protection and envelope awareness, respectively, as a spatio-haptic envelope. How this might be incorporated in surgical robotics is a challenge for both engineering and the psychophysics of human perception.
The latest generation da Vinci 5 robot has force feedback technology such that 43% less force may be applied to the tissue. This carries a risk of negative transfer if transitioning to a type without force feedback. In an aviation simulator study, when force feedback was introduced, it helped to guide pilots away from flight envelope limits and safety was improved on first use. However, if force feedback was subsequently removed, pilot performance degraded immediately. Similarly, a transition from a force feedback surgical robot to one without it could carry a risk of decreased performance.
Simulation
Simulation is ubiquitous in aviation in both initial and recurrent training. Flight simulation training devices (FSTD) are highly regulated in terms of fidelity of flight deck layout, environmental cues, and handling characteristics. This enables much flight training to be done in simulation. In robotic surgery, simulation is part of all training pathways. For console training (platform-specific), virtual reality (VR simulation) exercises establish competence at simple tasks. Beyond this, surgical simulation currently has limited fidelity. Therefore, speciality-specific training would still require cadaveric dissection.
Control and automation
In aviation, automation (‘autopilot’) is widespread. Larger aircraft are FBW. Converting control inputs into control surface movement (stick-to-control-surface relationship) is determined by a process known as ‘flight control law’. Manual flying in event of automation failure can be used in simulation scenarios. This remains a debate in aviation, given the role automation in several accidents. Though automation is not part of present-day robotic surgery, control plays a significant role. Using control law analogy, current surgical robots use ‘direct law’, with a direct hand-controller-to-instrument relationship. ‘Manual’ operating would be a conversion to laparoscopic or open surgery, with control law being ‘mechanical law’. Continuing the analogy, there is a similar debate in surgical training about reversion to more manual approaches. With the rise of robotic surgery, familiarity with laparoscopic and open approaches will decrease. However, it is clear in both aviation and surgery that precision of technical skills has increased with autopilot and robotic surgery.
Crew
In aviation, changes in flight deck technology and workload reduced cockpit crew through elimination of the navigator and flight engineer. This leaves only the two pilots, and speculation that increasing automation will create an even sparser flight deck in the future. In robotic surgery, role and location of theatre staff has also changed. The console surgeon sits remote from the patient and can manipulate 4 arms, which would in conventional laparoscopic surgery require two people. Bedside assistance is predominantly to manage the robotic arms. Robotic surgery has not experienced the decrease in surgical team size seen in aviation.
Cockpit and console
Pilot training starts with gaining the private pilot’s licence (PPL). The type rating for light aircraft is a broad category and includes a range of aircraft with largely similar cockpit layout and function. As aircraft become larger and more complex, the type rating contains fewer aircraft, such as a single aircraft (e.g. A380), or two aircraft designed with cockpit commonality to form a common type rating (e.g. Boeing 777 and 787). Similarly, in surgery, early stage training is in open and laparoscopic surgery. With this prior experience, robotic training has traditionally focused on using a particular robot type to perform an operation known to the learner. With less prior experience, there is a dual focus on both learning to use the robot as well as to perform the operation. With prior robotic experience, the concept of ‘transfer’ applies. Prior surgical and robotic experience therefore modulate the needs of a robotic curriculum into the following components: core (cross-platform, cross-specialty), specialty-specific, platform-specific, and platform-transitional.
Transfer
The concept of ‘transfer’ from the aviation industry helps explain this. Moving from one aircraft type to another can affect cockpit workload and performance through negative transfer or habit interference, conceptualised in a ‘transfer matrix’. The real-world significance of ‘transfer’ is highlighted by the several studies reporting a concurrent use of or serial transition from da Vinci. The impact of prior console expertise on performance has largely shown positive transfer. This is considered further below.
The team
Crew resource management (CRM) is formal mandatory team training used in aviation since the 1980s, focusing on non-technical skills. Non-technical skills training has also been delivered in surgery as the Non-Technical Skills in Surgery (NOTSS) course. Future iterations of this kind of course may bring a core (cross-platform) robotic element. Of note, the operating surgeon sits remote from the bedside, somewhat analogous to the radio-communication in aviation. Aviation communication tools such as standardised phraseology and read back may be particularly useful.
Human-machine interface
Dual control is standard in many aircraft; in general aviation, for training, and in commercial aircraft, to facilitate alternation of role. In robotic surgery, da Vinci offers a dual console option. Evidence is mixed regarding the benefit of dual over single console, noting loss of ‘embodied guidance’ with dual console. Other aspects of console and control layout can affect both platform-specific and platform-transitional components. Regarding console controls, da Vinci and Hugo have hand and foot pedal controllers, whilst Versius has only hand controllers. This increases hand activity in Versius users, but potentially reduces error in laterality of energy application because the energy control input is located on the same hand-controller that operates the instrument. This is an important difference, particularly relevant for a Versius-to-pedal-platform transition.
Another more general consideration is the lack of haptic feedback or force-feedback. In aviation, light aircraft have a direct mechanical coupling between cockpit controls and aerodynamic control surfaces. Force and vibro-tactile feedback in the cockpit is therefore an implicit property of the system. Modern airliners have flight-by-wire (FBW), which completely replaces the physical connection with electro-mechanical coupling. This is supplemented with flight envelope protection, which reduces pilot control inputs such that the aircraft remains within allowable limits. However, this envelope protection is not perceived by the pilot. This lack of feedback may have contributed to aviation accidents and has led to research in conveying flight envelope protection through haptic feedback in active control devices. By analogy, similar changes have occurred in surgery. Haptic feedback naturally occurs in open surgery, in direct proportion, and to a lesser extent in laparoscopic surgery, where forces are multiplied by the lever arm. Robot surgery is in effect surgery-by-wire (SBW). Robot surgery currently provides no haptic feedback, nor is there envelope protection. Though there is currently no clearly defined ‘surgical envelope’, dimensions of the envelope could include traction force and spatial reach, to constitute envelope protection and envelope awareness, respectively, as a spatio-haptic envelope. How this might be incorporated in surgical robotics is a challenge for both engineering and the psychophysics of human perception.
The latest generation da Vinci 5 robot has force feedback technology such that 43% less force may be applied to the tissue. This carries a risk of negative transfer if transitioning to a type without force feedback. In an aviation simulator study, when force feedback was introduced, it helped to guide pilots away from flight envelope limits and safety was improved on first use. However, if force feedback was subsequently removed, pilot performance degraded immediately. Similarly, a transition from a force feedback surgical robot to one without it could carry a risk of decreased performance.
Simulation
Simulation is ubiquitous in aviation in both initial and recurrent training. Flight simulation training devices (FSTD) are highly regulated in terms of fidelity of flight deck layout, environmental cues, and handling characteristics. This enables much flight training to be done in simulation. In robotic surgery, simulation is part of all training pathways. For console training (platform-specific), virtual reality (VR simulation) exercises establish competence at simple tasks. Beyond this, surgical simulation currently has limited fidelity. Therefore, speciality-specific training would still require cadaveric dissection.
Control and automation
In aviation, automation (‘autopilot’) is widespread. Larger aircraft are FBW. Converting control inputs into control surface movement (stick-to-control-surface relationship) is determined by a process known as ‘flight control law’. Manual flying in event of automation failure can be used in simulation scenarios. This remains a debate in aviation, given the role automation in several accidents. Though automation is not part of present-day robotic surgery, control plays a significant role. Using control law analogy, current surgical robots use ‘direct law’, with a direct hand-controller-to-instrument relationship. ‘Manual’ operating would be a conversion to laparoscopic or open surgery, with control law being ‘mechanical law’. Continuing the analogy, there is a similar debate in surgical training about reversion to more manual approaches. With the rise of robotic surgery, familiarity with laparoscopic and open approaches will decrease. However, it is clear in both aviation and surgery that precision of technical skills has increased with autopilot and robotic surgery.
Crew
In aviation, changes in flight deck technology and workload reduced cockpit crew through elimination of the navigator and flight engineer. This leaves only the two pilots, and speculation that increasing automation will create an even sparser flight deck in the future. In robotic surgery, role and location of theatre staff has also changed. The console surgeon sits remote from the patient and can manipulate 4 arms, which would in conventional laparoscopic surgery require two people. Bedside assistance is predominantly to manage the robotic arms. Robotic surgery has not experienced the decrease in surgical team size seen in aviation.
Author
Event Type
Oral Presentations
TimeWednesday, April 211:00am - 11:30am EDT
LocationPier 9
Simulation and Education (SE)