Research

Medical

Compliant magnetic devices can go places and reach autonomy that their more rigid mechanical cousins cannot. The M3Robotics lab is researching how to design and control highly flexible magnetic tools for both investigative and surgical use that can enable clinicians to have unprecedented reach and dexterity. Application areas range from neurosurgery to bronchoscopy.

Current Projects

Neurosurgical Electrode Guidance

neurosurgical electrode guidanceNon-invasive pharmacological treatments fail for some patients with brain-based neurological conditions, such as Parkinson’s disease, Tourette’s syndrome, epilepsy, and some forms of chronic pain. In these cases, the patient may opt for a neurosurgical solution. To address these conditions surgically, clinicians insert an electrode to stimulate a location deep within the patient’s brain.  Currently, there is no method to guide the electrode should it divert from the desired path. The goal of this project is to develop a system to actively and precisely guide an electrode to its target along any trajectory the clinician desires.

Sponsor: Boettcher Foundation

Past Projects

Cardiovascular Catheter Guidance

Cardiovascular Catheter GuidanceAtrial fibrillation is a condition that affects the ability of the heart to move blood efficiently.  To treat this condition, clinicians will ablate away areas within the heart to correct for abnormal neuro-signaling. Currently, this is performed using manually steered catheters with the patient and the doctor exposed to Fluoroscopic imaging X-Rays. The purpose of this project is to develop a closed-loop control method so that clinicians can guide ablation catheters precisely using magnetic fields.

Sponsor: European Research Council grant BOT-MED

Selected Publications

Magnetic

Magnetism enables engineers and scientists to effect objects without a mechanical connection. The M3Robotics Lab is researching how to design manipulation systems and use these systems to provide unprecedented control of both tethered and untethered objects. Projects range from fundamental investigations into materials to pushing the boundaries of what’s considered possible.

Mining

Underground environments offer many challenges to the men and women that work there. The M3Robotics Lab is researching how to bring autonomous platforms to reduce the risk to the workers and provide enhanced information to the mining companies. Projects range from autonomous navigation in underground environments to automated mineral detection.

Current Projects

Flying Underground

Drone in Edgar MineThe goal of this proof-of-concept project is to develop and demonstrate ground and air vehicles that are capable of autonomous navigation in the Edgar Experimental Mine. The primary focus will be on adapting a commercial UAV to autonomously fly in the confined space of an underground mine while building a 3D model of the mine as it travels and monitors key safety metrics such, as O2 levels. Such capability will provide many benefits to the safety and health of underground miners; for example, assessment of the risk of roof falls, the estimation of excavated volume, or assessment of ore pass blockage. Having an autonomous system perform these functions removes humans from these dangerous conditions. Once developed, this technology can be applied to other tasks such as inspection operations in open stopes, exploration of abandoned workings, and potentially search and rescue—where in the air vehicles search and ground vehicles rescue.

Sponsor: Alpha Foundation

NIOSH Smart Bit Project

Beam Acceleration response mode vs driving frequencyThe aim of this project is to develop a sensor framework that can be integrated into existing mining infrastructure to facilitate the estimation of rock strength parameters and ultimately, rock type. Sponsored by the National Institue of Occupational Safety and Health (NIOSH), this project will help mitigate the risk of roof collapse by alerting operators to changing mining conditions. Currently, this task is performed by experienced operators using audio-visual cues which is both difficult and subjective. Providing assistance to this operators in the form of objective measurements obtained from the machine/material interface will no doubt improve both safety and efficiency of these mines. This project will leverage analytical digging models developed as early as the 1960’s as well as modern Machine Learning techniques coupled with signal processing insight. This technology will be built on a general platform such that it can be adapted to many mining machines and force sensing applications.

Other

Current Projects

Power Plant Waterwall Inspection & Repair Robot

Sponsors:
             

In this project, an integrated robot is developed, equipped with nondestructive sensors to be used for live inspection. It operates an onboard repair device which performs live repair and utilizes artificial intelligence for data fusion and live predictive analysis. All features of the robot are for the purpose of automated spatiotemporal inspection and the repair of furnace walls in coal-fired boilers. Enabled by artificial intelligence, this robot can automate data gathering such as 3D mapping, as well as live predictive analysis, then learn from user feedback, continuously improving performance and achieving smart autonomy. The project’s success would have significant benefits, through decreasing the need for operators to access dangerous or difficult areas, eliminating time-consuming scaffolding, enabling automated inspection and repair, and intelligently gathering comprehensive data. This project will have a tremendous impact, saving cost and time, reducing risk to human operators, and increasing boiler productivity and reliability. 

                     

Fall 2020a new CAD model was designed for the gantry system, with ball screws and railed carriages, which made the system safer towards dynamic loading. Electromagnets were also installed, which allows the system to ground itself to the climbing surface while stir-welding. The system was redesigned so that NDE sensors can be mounted about the bottom lateral strut, which allows for rescanning and retained accuracy when the robot drives. What’s more, a geared brushed motor was chosen, whose lower voltage and amperage allows the Arduino and Adafruit motor shield to control the entire gantry system autonomously. With this progress, the gantry system is made safer, can mount to the surface it’s climbing, can rescan the surface to ensure accuracy in its special awareness, and can be controlled by Arduino code. 

Winter 2020-21, all the load bearing machined components were tested, and passed their required Factor of Safety, making the overall system far safer. As well, as a study was performed to confirm that the robot is capable to detecting cracks, and it was determined that the sensors can even detect cracks despite lift-off distance from the surface, in a range of 0-5mm. This is especially important, as the sensors will be able to find cracks without having to stay entirely flush with the surface until the actual stir-welding process using electromagnets. 

Spring 2021, the weight of the entire system was considered, and stabilizing 8020 bracing, as well as hoist rings and handles, were added to assist with lifting. Now, the system is manageable to be lifted where it is needed. An emergency stop was added that can immediately shut down the system, making it so the system can easily be shut down if needed, avoiding damage to itself or the surface it is welding. As well, cooling fans were added, so the electronics will be protected from overheating and damage and can thus work effectively as intended. 

Summer 2021, progress was made in the code, so that a closed loop code from the laptop can issue commands to the drill. The drill can move along the X and Y coordinates, and the electromagnets now work to attach and detach. Finally, sensors were wired and installed to be able to detect the edges of all directional planes, ensuring the drill does not move off the rails, or past where it needs to be, at risk of damaging the system or code. 

The system can now latch onto the surface it is stir-welding, upon which it will need to be secured for the stir-welding process. The system now has many safety features installed to ensure the system does not break, the boilers it is working upon are not damaged, and any people working with the system can do so easily and safely. The motors work, meaning the drill can move now, across both the X and Y axes, and can even move across both simultaneously. It has many sensors to scan the surface for cracks, as well as internally to determine the range along which the drill can safely move. 

Robotic Motion and Calibration for Antenna Characterization

Antennas are a part of everyday life, allowing for the fast and convenient transfer of information in many applications. Understanding the properties of each antenna as manufactured requires high accuracy readings across a large number of volumetric points. While the positioning of sensors during this process has traditionally been accomplished with specialized rail systems, a transition is happening in the measurement community to robotic positioning systems to introduce versatility in pose selection and operation range. The Colorado School of Mines is collaborating with NIST in order to enhance the accuracy and motion planning capabilities of such systems with a focus on collaborative robot motion.

Sponsor: NIST

Soft Robotics Using Shape Memory Materials

The research is focused on modeling and 3D printing composites of Polyurethane (PU) based shape memory polymer (SMP) and NiTi shape memory alloy (SMA) wires for soft robotic applications using COMSOL Multiphysics. The composite is “trained” to undergo bending deformation upon Joule heating actuation. An inchworm inspired motion is mimicked to convert the bending deformations as useful locomotion. The thermo-mechanical constitutive model is implemented in COMSOL Multiphysics to study the complex deformation based on temperature and heat generation. The actuator design is optimized for Cost of transport (COT) using the Derivative-free Monte-Carlo method inbuilt in COMSOL Optimization module.

 

Selected Publications