Assessment of hand kinematics is important when evaluating hand functioning. Major drawbacks of current sensing glove systems are lack of rotational observability in particular directions, labour intensive calibration methods which are sensitive to wear and lack of an absolute hand orientation estimate.

Emerging microelectromechanical system (MEMS)-based sensors become much more applicable for on-body measurement purposes lately. Especially, the development of a finger tip-sized tri-axial force sensor gives the opportunity to measure interaction forces between the human hand and environmental objects. We have developed a new prototype device that allows simultaneous 3D force and movement measurements at the finger and thumb tips. The combination of interaction forces and movements makes it possible to identify the dynamical characteristics of the object being handled by the hand. With this device attached to the hand, a subject manipulated mass and spring objects under varying conditions. We were able to identify and estimate the weight of two physical mass objects (0.44 kg: $29.3\%\pm 18.9\%$ and 0.28 kg: $19.7\%\pm 10.6\%$ ) and the spring constant of a physical spring object ( $16.3\%\pm 12.6\%$ ). The system is a first attempt to quantify the interactions of the hand with the environment and has many potential applications in rehabilitation, ergonomics and sports.

In this paper a new concept to identify environmental loads during the interaction with the human body by sensing interface forces and movement is proposed. Mass and spring loads were moved by hand over a fixed height difference. Kinematic and kinetic quantities were measured between the hand and the load using an instrumented handle. Force was measured using a force transducer module, movement was measured using an accelerometer and rate gyroscope. Under the condition that the human body was actively generating force at the load, while the load was passive, the dynamic characteristics of the load could be estimated. The estimated parameter values were compared to their specified values and appeared to be accurate within 4% for both mass and spring loads.