What Is Force Control?

Force control is a branch of control systems engineering concerned with regulating the forces and torques that a mechanism exerts on its environment, rather than specifying only its position or velocity. In tasks such as robotic assembly, polishing, grinding, or surgical tool guidance, pure position control is insufficient because contact forces depend on the stiffness of the environment and small position errors can produce large, damaging forces. Force control addresses this by using force and torque measurements as feedback signals to modulate actuator outputs and maintain desired contact conditions. It draws on classical control theory, rigid-body mechanics, and sensor technology, and it is a central topic in robotics, manufacturing automation, and human-machine interaction.

The need for force control arises whenever a robot or mechatronic system must interact physically with objects, surfaces, or humans. Unlike free-space motion, contact tasks impose kinematic constraints on the system, and the control law must account for both the desired motion directions and the desired force directions simultaneously.

Hybrid Position-Force Control

Hybrid position-force control, introduced by Mason and formalized by Raibert and Craig in the early 1980s, partitions the task space into orthogonal subspaces: directions in which position is controlled and directions in which force is controlled. In a peg-in-hole insertion, for example, the robot controls force along the insertion axis to avoid jamming and controls position in the lateral directions to maintain alignment. The control architecture uses a selection matrix to route position and force references to the appropriate feedback loops. ScienceDirect's overview of hybrid position-force control covers the theoretical foundation and outlines the practical extensions developed for non-rigid environments and redundant manipulators.

Impedance and Admittance Control

Impedance control, developed by Hogan at MIT in 1985, takes a different approach: rather than specifying a desired force directly, it specifies a desired dynamic relationship between end-effector displacement and contact force, modeled as a mechanical impedance (mass, damping, and stiffness). The robot is made to behave as though it were a spring-damper system with programmable parameters, which produces naturally compliant contact behavior without requiring an explicit environment model. Admittance control is the dual formulation: force measurements drive a position command rather than the inverse. Admittance control is more suitable for stiff mechanical structures common in industrial robots, while impedance control is often preferred in light-weight or cable-driven systems. IEEE Xplore research on unified impedance and admittance control explores hybrid formulations that combine both strategies to handle environments with varying stiffness.

Force Sensing and Implementation

Effective force control depends on accurate, low-latency force measurement. Six-axis force-torque sensors mounted at the robot wrist are the most common transducer; they use strain gauges arranged in a Wheatstone bridge configuration to measure three force components and three torque components simultaneously. Joint torque sensing, used in torque-controlled robots such as the KUKA iiwa and the DLR LWR series, provides an alternative by estimating contact forces from joint torque readings, avoiding the expense of a wrist sensor at the cost of reduced accuracy in the presence of friction and gear elasticity. Closed-loop force control typically requires sensor update rates of 1 kHz or faster to maintain stability at contact, and control law computation must complete within the same cycle. Research on robust hybrid impedance control published through IEEE Xplore documents early stability analysis methods that remain foundational to current implementations.

Applications

Force control has applications across a wide range of fields, including:

• Robotic assembly of tight-tolerance parts in automotive and electronics manufacturing
• Surgical robotics and minimally invasive tool guidance
• Physical rehabilitation exoskeletons providing adaptive resistance
• Grinding, polishing, and deburring automation for surface finishing
• Human-robot collaboration where safe compliant interaction is required