The principles of fracture surgery are accurate alignment of bone fragments, rigid fixation, avoidance of soft tissue injury, and early recovery. Generally, it has priority to restore a bony structure between a proximal bone fragment and a distal bone fragment than to fit all pieces of bone fragment of diaphysis. A repositioning to normal anatomical structure is important in upper limbs fracture, whereas a restoration of mechanical axis is emphasized in lower limbs fracture. Bone fragments should be rigidly fixed not to make movement at the fracture site.

There are two main methods for fixation of bone fragments; a closed intramedullary nailing and a plate osteosynthesis. The closed intramedullary nailing is now considered as a standard treatment for a long bone shaft fracture. The plate osteosynthesis has showed particularly advantageous when an intramedullary nail may be technically not feasible.

Lee et al. reported that MIPO (Minimally Invasive Plate Osteosynthesis) had the shorter bony union time and the longer operation time than interlocking intramedullary nailing. Two groups did not show statistical significance in clinical results [24]. Apivatthakakul et al. [2] reported that although the biomechanics of the plate fixation are less stable compared to the intramedullary nail, the mechanical stability is stable enough for bone healing. Guo et al. [13] conclude that both an intramedullary nailing and a percutaneous locked compression plate can be used safely to treat distal metaphyseal fractures of the tibia from a prospective comparison study. They prefer the intramedullary nailing, because of shorter operating and radiation time, and easy removal of the implant.

If surgeon makes incision at fracture site for accurate bone alignment and rigid fixation, this incision may be cause of a soft tissue injury, a failure or delay of a bone healing, and an infection. Over the past decades surgeons have tried to fix bone fragments by inserting implant away from the fracture site through minimally invasive incision. This surgical technique, called minimally invasive fracture surgery, can keep blood flow to the injured tissues and reduce the risk of infection. Consequentially, it shows good surgical results and early recovery than the conventional surgical methods.

In early state of MIPO (minimally invasive plate osteosynthesis), David et al. were confident that the MIPO technique for the treatment of distal tibial fractures would be a feasible and worthwhile while avoiding the severe complications [14].

Mahmood et al. [29] reports that minimally invasive technique shows less blood loss, minimal soft tissue destruction, shorter hospital stay, and early mobilization comparing to conventional methods in dynamic hip screws for fixation of intertrochanteric fractures of femur. Wong et al. [46] concluded that MIDHS (Minimally invasive Dynamic Hip Screw) fixation for intertrochanteric femoral fractures is superior to the conventional technique from their double-blind, prospective, randomized, and controlled clinical trial. The MIDHS produces less blood loss, less pain and a shorter rehabilitation period, while still achieving good radiological outcome.

MIPO for mid-distal humeral shaft fractures could effectively treat with advantages of shorter fracture union time and lower incidence of iatrogenic radial nerve palsies but with similar functional outcomes to the conventional open plating technique [1].

The robot system has two main roles in fracture surgery; one is a guidance or insertion of a needle or a screw, and the other is a positioning the bone fragment for a fracture reduction.

Though there are already some commercialized surgical navigations for determining the insertion positon, only a few laboratory-level’s robotized systems are reported. K. Bouazza-Marouf et al. first introduced robot assisted orthopedic surgery. Their robot allowed a drill-bit guide to be automatically aligned with an intra-operatively planed drilling trajectory [5]. Shoham et al. [41] have developed a bone-mounted miniature robot to precisely position and orient a drill or a needle for the same use.

The robotic system for assisting the fracture reduction is also not commercially available until now. A reduction robot system called “RepoRobo” was firstly introduced by Fuchtmeier et al. [9]. They converted a commercial industrial robot for medical use by appropriate modification. The requirements for using the industrial robot for the reduction of femoral shaft fractures are well described.

Westphal et al. also tried to use an industrial robot for medical use. They showed that robot assisted fracture reduction for the femur provides a high precision in alignment, while reducing the amount of intra-operative imaging from the statistical analysis [11]. Then, they have developed a surgical telemanipulator system to support long bone fracture reduction procedures [44, 45]. A joystick with force feedback was installed in their system as the control device for the telemanipulator, and the interaction for 3D fraction reduction with a 2D joystick input device was described. They showed the very accurate reduction results in their experiments with the telemanipulator.

Gramham et al. introduced a parallel robot for long bone fracture reduction [12, 31]. They used a foot holster to fix the leg to the parallel type robot. The geometric model of the bone fragments were visualized in a monitor. They also introduced force modeling during fracture reduction to determine the requirements for a robotic device.

Ye et al. [47] suggested a hybrid robotic system to balance the accuracy, payload and workspace. And Tang et al. [42] showed a hexapod computer-assisted fracture reduction system could reduce long-bone diaphyseal fractures effectively.

Mitsuish et al. had developed a fracture reduction assisting robotic system, which consists of a newly designed robot for only medical user, and a reduction path navigation system based on 3D CT imaging [30, 43]. The foot is fixed using a boot like the conventional fracture table. Clinical data of reduction forces/torques are reported with this robotic system by Maeda et al. [28] as mentioned before. An automatic reduction for fracture surgery was also developed by modifying this system [16]. The robot was fixed to a distal bone fragment through two fixation pins and the goal position and the reduction path were generated from a navigation system based on surgeon’s opinion. The phantom experiment results were very promising.

Fuchtmeier et al. [9] suggested some safety features of “Reporobo”. First, the robotic system has the velocity limits using the large reduction gear ratios, though the general operating speed of the robot may be controlled via software. At the same time, “back-driven” should be possible so that the surgeon can physically move the robot arm away from the patient in an emergency. “Harmonic drives” were applied to their system in this end. Secondly, they used a robotic gripper with integrated compliance to prevent the robot from pulling the fixator out. Lastly, they designed the control software for the force sensor to enable a free definition of the load limits to avoid over-correction and over-extension of the extremity.

Westpal et al. [44] took safety precautions by dividing the project into two layers, a software layer and a hardware layer, while developing the surgical telemanipulator. The software layer checks all the variable sensor information against predefined thresholds. If one safety threshold is exceeded, a stop command is sent to the robot control unit as soon as possible. For instance, if forces/torques exceeds a defined threshold, the motion of the robot is immediately stopped, and subsequently only those motions that lead to a reduction of the applied forces are allowed. Position information is used to limit the translational or rotational speeds. The hardware layer also limits the forces or torques with a load limiter (ULS 100, IPR-Intelligente Peripherie fur Roboter GmBh, Schwaiger, Germany). As soon as the forces or torques applied to this device the thresholds, the robot controller performs an emergency stop.

Gramham et al. [12] explained that parallel robots have inherent safety such as an increased stiffness, a high gear ratio, and high accuracy. Thus, they used the parallel robot mechanism for fracture reduction. And it was said that they plan to incorporate safety features such as a watchdog timer, a dead-man switch, force monitoring, encoder redundancy, and software motion limits.

Here, safety and system design methods will be described based on developing experience at the University of Tokyo [17]. First of all, inherent hazards of the surgery and the robot-induced hazard should be listed up. Figure 2 shows the hazards related to a robot-assisted fracture reduction. An indirect reduction means that connection part does not directly connect to the bone fragment. In this way, limbs are just grabbed by hands or seized with a connection part like a boot. The conventional surgical technique belongs to the indirect reduction. A direct reduction method was suggested to improve reduction accuracy by fixing an end effector to a bone fragment using fixation screws. And a robotic system was introduced for more accurate and safe fracture reduction. Though the accuracy of the reduction can be increased by applying the new technique to the conventional methods, the extended system also results in an increase of number of hazards. However, if the robotic system is able to control the hazards satisfactorily, the application of this system could increase reduction accuracy with lower risk. Some hazards that must be controlled and its preventive measures will be suggested.