Mechanism Design

Design of a low-­cost and easy to use robotic vision-­guided micromanipulation platform


At present, micromanipulation relies a lot on manual control. This makes experimental outcomes highly dependent on user’s skill and experience. It not only limits the speed and ease of the operation but also affects the reproducibility of a study as well. The need for consistent, systematic, and efficient micromanipulation motivates development efforts for robot-assisted microscopy and manipulation. Despite the advancements in robotic vision-guided technology, its potential is not fully tapped due to the challenges in setting up and operating robotic vision alongside with existing microscopic imaging system.

In this work, a robotic vision-guided platform designed for low-cost deployment and easy operation is proposed to address the identified need. The goal is to develop an uncalibrated, self-initialized, and auto-targeting robotic vision-guided control. The advantage over existing systems is its self-contained operation workflow that requires no complicated setup and manual intervention.

It is envisioned that this work will eventually contribute towards the ease of implementing robotic vision-guided micromanipulator platform. The long-term goal is to open up new possibilities in the ways experimental biology is being carried out with our low cost and easy to use platform design.

Partner: Professor Kamal Youcef-Toumi (MIT)

Researchers involved: Liangjing Yang, Ishara Paranawithana

Gravity Compensation for Easy Loading/Unloading of Heavy Loads from Truck

A greying population is a global phenomenon, especially in Asian countries like Japan, China, Singapore, South Korea and many countries in Europe. Along with long-term decline in birth rates, the shrinking proportion of working-age people (ages 15 to 64) raises serious economic concerns. Countries like Japan and Singapore with total fertility rates (TFR) way below replacement level fertility of 2.1, turn to technologies to add value to their workforce through improving work efficiency in labour-intensive occupations (e.g. military, construction and farming etc.). One of such technologies are gravity compensation devices. In Japan, companies like Cyberdyne Inc. created exoskeletons that reduces load’s effective weight and companies like Obayashi Corp. are adopting them to extend the working life of elderly workers.

This research project aims to develop a novel passive gravity compensation device for loading and unloading of heavy loads from vehicles. Existing practices use actively powered hydraulics and pulley systems which are limited in its range of motion, bulky and inefficient for medium loads. Deployable mechanical ramps and trolleys are not only time-consuming but possess occupational safety risks. Using design science methodologies, the project aims to leverage on existing designs and draw inspiration from other fields to develop a multi-DOF lifting mechanism that is easily deployable and able to accommodate to a large range of workspace and loads. Potential applications of such a design can be in parcel delivery services and in military settings where increased work place safety and more efficient use of limited manpower will be appreciated.

SUTD & IDC Faculty involved: Professor Kristin L. Wood

Researcher involved: Dexter Chew

Algorithm for Designing Mechanisms with Non-Linear Negative Stiffness Property

Negative stiffness mechanisms have been widely used in various mechatronics applications such as vibration control, compliant control, and robotic grasping. Unlike the normal springs where the magnitude of the reaction force increases with deformation, negative stiffness compliant mechanisms provide a decreasing force/torque magnitude with deformation and are used in many applications. This paper proposes an algorithm that aids users in designing mechanisms with negative stiffness property.

By determining a prescribed stiffness function that meets the requirement of the application, the proposed algorithm is able to automatically generate an optimal compliant beam design to match the prescribed stiffness function in both translational and rotational applications. Two examples, namely one in rotational perspective and the second one in translational perspective, are used to demonstrate the capability and efficacy of the proposed algorithm. Experiments are conducted and practical applications which include: a) adjustable constant force gripper; and b) multi-DOF gravity compensation for upper limb are illustrated.


SUTD Faculty involved: Shaohui Foong

Researcher involved: Zhuoqi Cheng