Research | MAAL

Sensorless contact wrench estimation for industrial robots

Tue, 11 Aug 2026 00:00:00 +0000

Robots used for polishing, assembly, and manipulation need reliable contact-force feedback. A six-axis wrist force/torque sensor provides direct measurements, but adds cost, wiring, payload, and integration constraints. This project estimates the same six-dimensional contact wrench from signals already available on the robot: joint motion and motor current.

Dual-stage estimation

The estimator combines a physics-based robot model with two learned residual corrections:

Free-space dynamics compensation. An LSTM learns joint-torque errors caused by friction, backlash, hysteresis, parameter mismatch, and torque-conversion bias using contact-free trajectories.
Contact wrench compensation. A temporal encoder and residual network correct the remaining task-space error during contact and predict an input-dependent uncertainty for each wrench axis.

Separating these responsibilities keeps the estimate anchored to robot mechanics while allowing each learning stage to focus on a distinct error source.

Experimental validation

The method was evaluated on a six-axis industrial manipulator. A cable-and-pulley rig applied loads of 0.5, 1.0, and 1.5 kg from multiple directions. A wrist load cell sampled at 1 kHz supplied training labels and evaluation ground truth, but was not used as an input during sensorless inference.

Stage I reduced the Joint 1 residual-torque RMSE from 4.2649 to 0.8591 Nm and the Joint 2 RMSE from 3.1453 to 1.0069 Nm. Across the final six-axis wrench evaluation, the dual-stage method achieved the lowest reported RMSE, maximum error, and relative error on every force and moment axis.

Metric	Fx	Fy	Fz	Mx	My	Mz
RMSE	1.5959 N	0.7067 N	1.9678 N	0.1367 Nm	0.1141 Nm	0.0538 Nm
Maximum error	5.9546 N	4.4262 N	5.3414 N	0.7739 Nm	0.4898 Nm	0.4104 Nm
Relative error	13.8%	5.8%	13.4%	17.6%	22.2%	8.3%

Why it matters

The main contribution is not simply a deeper network, but a physically meaningful decomposition of the estimation problem. The intermediate output makes each correction inspectable, while the probabilistic second stage provides a condition-dependent confidence signal. This creates a practical path toward lower-cost force-aware robots and future closed-loop force control without a permanently installed wrist sensor.

Flexible industrial robot dynamics and control

Tue, 06 Jan 2026 00:00:00 +0000

Industrial robots change stiffness and vibration behavior with their position and posture. This project establishes dynamic models that capture both joint and link flexibility without making the model too costly to use in control.

Approach

Represent geometrically nonlinear links with strain-based beam elements and modal coordinates.
Identify configuration-dependent vibration from experimental data.
Generate nonlinear filtered B-spline trajectories that suppress motion-induced vibration.
Validate compensation strategies on industrial robotic platforms.

Why it matters

More accurate dynamic models allow robots to move faster while maintaining precision. The same methods support robotic machining, additive manufacturing, and large-reach manipulation.

Additive manufacturing process monitoring and control

Mon, 05 Jan 2026 00:00:00 +0000

This project combines additive-manufacturing process models with the dynamics of the motion platform. Thermal cameras, nozzle cameras, accelerometers, and position sensors provide synchronized data for monitoring and compensation.

Research questions

How can process and machine dynamics be represented in one control-oriented model?
Which visual and thermal signals reveal defects early enough for correction?
How should local and cloud controllers share model-based and data-driven tasks?

Expected outcomes

The project develops in-situ monitoring, extrusion-trajectory compensation, and control strategies that improve dimensional accuracy and process stability.

Continuous-fiber additive manufacturing

Sun, 04 Jan 2026 00:00:00 +0000

Continuous fibers can make additively manufactured components significantly stronger and stiffer, but only when the fiber paths follow the structural demands of the part.

Research scope

Build computationally efficient models of free-form composite parts.
Optimize fiber paths for modulus, strength, mass, cost, and manufacturability.
Translate optimized paths into reliable multi-axis fabrication.
Measure the printed parts and connect observed performance back to the model.

The result is an integrated design-to-fabrication workflow for high-performance composite components.

Flexible aircraft reduced-order modeling and load alleviation

Sat, 03 Jan 2026 00:00:00 +0000

Flexible and very flexible aircraft improve structural weight and aerodynamic efficiency, but their flight dynamics are tightly coupled with structural deformation.

Approach

Construct reduced-order structural and aerodynamic models.
Preserve the aeroelastic behavior needed for control design.
Use the model to evaluate load, fuel efficiency, and flight trajectories.
Develop load-alleviation and flight-control algorithms.
Verify models and controllers through wind-tunnel experiments.

The project makes high-fidelity aeroelastic behavior accessible to analysis and real-time control.

VTOL data-driven modeling and control allocation

Fri, 02 Jan 2026 00:00:00 +0000

Vertical takeoff and landing aircraft operate across flight regimes where aerodynamics, rotor-wing interaction, and actuator authority change substantially.

Approach

Combine physics-based models with flight-test data.
Identify aerodynamic effects at high angles of attack.
Model rotor-wing-airframe interaction and thruster vectoring.
Use the hybrid model in a control-allocation framework across takeoff, transition, cruise, and landing.

The aim is a model that is both physically interpretable and accurate enough for robust flight control.

Boom-lift-mounted robotic systems

Thu, 01 Jan 2026 00:00:00 +0000

Mounting a robotic manipulator on a boom lift extends its workspace, but the flexible, extendable support introduces configuration-dependent vibration and positioning error.

This project models the coupled boom-and-robot system, identifies changing structural dynamics, and develops trajectory and active-compensation methods for accurate long-reach operations.

Applications include inspection, maintenance, construction, and manufacturing tasks where conventional fixed-base robots cannot reach.