Meeting Minutes: Microrobotics
April 15, 2017
April 15th 2017
Jianxiong Wang: This friday night I was working on the ROI of the robot body and I used RANSAC method to improve the quality of feature detections. To obtain a rotated ROI, I identify the center of the robot and its orientation. With the rotated ROI, I reset the orientation of the robot to cancel out the rotation in the actual scene so the "Rotated Image" (the small window in the figure) is always vertical. Ideally, the robot in the rotated image should always have the same orientation that the object template has, but my orientation calculation has about 20 degrees of error. I will keep finding solutions to increase the accuracy of the calculations. With the rotated image, we can perform more detailed templating algorithm.
February 2017
February 14, 2017
Jermaine:
I added two potentiometers to the fatigue-test apparatus. I have been developing a code in Arduino to utilize both devices. Right now, the pots are having trouble manipulating servomotors. This week I will continue figuring out the cause behind this issue.
Bo:
I calibrated the operation time of motor for each locations at x direction. Based on the calibration results, I modified the code in Arduino by adding an increments on the operation time. Right now, the x-direction movement move much smoothly than before. Next step is to decrease the resolution of the x-directional movement.The other two directional movements are very good. In addition, I plan to build the holder in the solidworks and test the manipulator with camera.
Ze:
Jianxiong:
I modified my existing code for the video tracking so that it can accomodate the new method. I am still learning the implementation and the concept of transformation matrix. This week I will be trying to implement the transformation matrix.
Maria:
Benji:
Vinoth:
February 7, 2017
Jermaine:
Updated all 12 types of nerve retractor models in and pulley designs in Solidworks. Identification codes were added to each part to distinguish nerve retractor types. Final models were sent out to be printed.
As of now I am working on incorporating a potentiometer to the fatigue-test apparatus. This will enable better control as well as minimizing damage to the nerve retractors during calibration of the system. The code will eventually be added to the current fatigue-test code.
Bo:
I refreshed and reprogrammed the new ESC successfully. I assemblied the reprogrammed ESC to the Y direction of the manipulator. In addition, I added some codes in MATLAB and Arduino to control the Y direction movement of manipulator. Now, the manipulator can move in three dimensions.
In order to make sure the manipulator move smoothly, I still need to do some tests to caliberate the throttle signals and operation time of brushless motor at different locations of the manipulator.
Ze:
Jianxiong:
Here are the things that we should pay attention regarding the tracking of robot.
To have better tracking performance, the body of the robot should have less light. Due to the surface of the robot is reflective, any change on the orientation of the robot can result in a significant change on the lightening condition which can result in the loss of many feature points. Once the light condition is stable, transformation matrix could be applied to find out the orientation of the robot body first. With the orientation of the robot, we could find the approximate location of the probe. Then, region of interest is applied to the area around the probe to do a template matching. As you can see from the picture below, the current problem is still that the probe can only have two or three feature point, but such problem could be solved with our new design (bigger probe).
In general, the new system has more pixel points than the old video. The same method was applied to the old system, the one that Wuming did, but the result did not go well. There are two methods that we could use for the new design.
1. Hough Transform (simpler method)
a. This one strictly depends on the geometry of the probe
2. Transformation matrix + template matching
a. This one is not strictly dependent of the geometric shape of the probe as long as the probe has enough feature points
Before our new system comes out, I will be working on the second approach, and I will work with Maria and Ze to shoot a new video with dyed background.
Vinoth:
I was able to calibrate the manipulator and fine tune it with respect to the camera frame. I have also included an automatic calibration protocol in the GUI by allowing users to click points on the manipulator tip to perform the calibration. After the points are collected, a simple linear regression is used to determine the rotation angle and offsets.
Tested the multiple point soldering with real time vision feedback.
Benji:
1. Continued working on the microrobot swarms hardware for Duke experiments. Created a new GUI for easier troubleshooting of coils.
2. Conducted experiments with the nanowires with Maria on the Magnebotix system.
Maria:
· µFSSM: (1) First photolithography steps with the new mark (with the new designs), the patterns look well defined and the transparencies seems to give a good resolutions. (2) I did the training for the back alignment and the wafer is ready for the etching tomorrow (I’ll follow the steps Wuming described in his mail. I'll do the release with this sample and an old one of Wuming.
· Experiments with the cells: We checked out the cells under the microscope and the microbots. We prepared some microrobots with Ze in the weekend but they broke when we tried to cut the magnet.
· Experiments with the NW: We did the first batch of experiments yesterday to check out if the parafilm channel thickness could help on the visualization (it doesn’t). We tested with Ze and Ben two more samples of NW and although its motion is appreciable in the x-y plane, in y-z is not that clear. By Friday we had checked them all and we will complement the report and upload it. Hopefully with the Ni flakes we will get some more results.
· Postdoctoral Challenge: I’m preparing and putting in order the ideas and I discussed with Perry Kirkham. He told me to talk to Jaqueline Linner to check out the project, as she is working with point of care devices and she recently a proposal in NIH (the grant will be evaluated under the same criteria as the NIH proposals).
May 2016
May 24, 2016
Wuming:
- Have demonstrated the calibration test (setting and process) to Benji.
- Have upload the files to the group server, that related to my past microrobot research; and the document for the notes of our micro-fab supplies. They are in folder " ~/MSRAL/Current Projects/Microrobotics/Wuming/ ", the sub-folders and files are:
DesignSimulation/ (all the mask and COMSOL files)
Dissertation/ (the files about my PhD dissertation)
Literatures/ (the literatures looked up about past microrobot research)
Papers/ (the files for all our submitted publications)
Results(PicVideoEtc)/ (the fab, test photos and videos)
zSuppliesMemo_wj.pptx (the powerpoint file documenting the used micro-fab supplies used in the past)
May 3, 2016
Wuming:
Research on dictyostelium
The reported traction stress for investigated dictyostelium is on the magnitude of 1E3 dynes/cm2, i.e., 5E3 dynes/cm2 = 5E-4 (µN/µm2). The typical size of dictyostelium cell is ~ 10 µm, which is kind of too small for our current µFSMM. Furthermore, assuming the cell occupies the contact area as 10 µm square ~100 µm2, times the traction stress, 5E-4 µN/µm2, we can have the typical force is on the order of ~ 0.05 µN = 50 nN, which is also some small for our current prototype.
Have finished part of the slides of gripper (macro-scale), for IDETC 2016
April 2016
April 26, 2016
Wuming:
Have fabricated new field/arena for MMC/Bo.
Have located all the wares/supplies at Birck, in and out the clean room. Will document them around Mid-May.
For Dr. Maria: have drawn series of .stl files of different minimum key dimensions. Waiting for their feedback of the fab trial to follow up.
April 19, 2016
Wuming:
Model & FEA the micro-gripper design for Nanoscribe
See the inserted figure:
The footprint < 200 µm. The beam width = 8 µm. The device thickness = 15 µm. The material Young's modulus E = 2 GPa, Poisson ratio = 0.35. The gripper gap = 56 µm wide.
---> In order to close the gap, the contraction force (indicated by the black arrow) required = 14 mN, while the maximum tensile stress that Nafion@, SMP material can handle is around 20-30 MPa. Therefore, assume using Nafion@, where the contraction stress is 15 MPa, the required diameter of the Nafion wire is approximately 35 µm in diameter, or equally cross-section area.
Will make a few more wafers of assembly field for MMC/Bo, this week.
Bo:
have finished with ME 497 poster and summary.
have done the assembly challenge with four triangular components, and recorded some test videos.
April 12, 2016
Bo:
have tested with the new thin triangles and recorded some test videos.
simplified the Arduino codes.
April 5, 2016
Bo:
continued to practice with the assembly challenge and recorded some test videos.
continued to practice with taking down the new triangles from silicon chip.
Ze:
Tracking works very well with the new silicon chip embedded chamber .
Start rewriting the control part. Decided not using PID controller, which always gives a much too high gain when robot gets stuck.
March 2016
March 28, 2016
Wuming:
Revision on the ufsmm paper.
Modified the ufsmm design for nanoscribe capability (design with 5 um thick spring beam, footprint < 250 um ). The FEA model in COMSOL shows that stiffness far from desired --> Material property dominates the stiffness, same like our real test with the previous PDMS. Simulation here has plugged the SU8, Young's modulus E = 2.0 GPa, while the previous generation ufsmm uses normal PDMS's modulus at 0.5 - 1.0 MPa.
Fabricated thinner triangle pegs for MMC.
Bo:
continued to practice the assembly challenge, and recorded some videos.
have tested the new arena made by Ethan.
have tried the smaller and thinner triangles and recorded some videos.
March 21 2016
Bo and Ze:
Laser cutting the arena holders .
Bo has practiced with the assembly challenge with the new top coil cylinder and made some videos.
March 15 2016
Bo and Ze:
Have found the hole on the printed top cylinder is too small, the transmission of light is poor. we have built a new top coil cylinder with some changes in the Solid Works.
have built the model of boundary for arena in Solid Works and been ready for the laser cutting.
have wrapped the coil around the top coil cylinder.
Wuming:
Drafted the uFSMM journal paper.
March 8 2016
Bo and Ze:
Build the 3-D model of top cylinder in Solidworks, and use 3-D printer to print it out.
Bo is trying to modify the assembly codes to reduce the time delay from 200 milliseconds to 50 milliseconds, so the robot can respond faster.
February 2016
February 29, 2016
Bo Ji and Ze An:
- we were trying to add another coil cylinder above the device to reduce the friction on the robot. it worked very well.
-Bo made two assembly challenge videos by using new added coil cylinder.
February 23, 2016
Wuming:
- Writing on book chapter, 13 pages so far.
February 16 2016
Bo Ji and Ze An
- can change the direction of current in the coil cylinder, so the coil can generate the repulsive force, but after test with the robot, it seems the repulsive force doesn't work very well.
January 2016
February 9
Bo Ji:
have tested code on the competition device, and modified the codes to make robot move smoothly on the competition device.
already can set everything all by myself.
have made two videos of recording the movement of robot in the competition device.
Bo Ji:
have tested four models of speed by using Sagar's robots with Ethan and An Ze and made some videos.
have learned and practiced how to set up the whole device with the help from Wuming.
have used Solidworks to simulate the shapes of robots.
January 26, 2016
Bo Ji:
Replacing the acceleration trigger with four buttons which can switch the speed modes. Consider robots may experience different conditions, different speeds are required to make to the robot move more smoothly and avoid the stuck situations. I added four buttons(A,B,X,Y) form Xbox controller to represent different speeds. The choice of speed depends on the situations robots experience.
Button X sets the PWM to 540;
Button Y sets the PWM to 580;
Button B sets the PWM to 620;
Button A sets the PWM to 650;
Wuming:
- MARSS 2016:
have modeled the micro-scale multi coil testbed. Data have been taken for Sagar, for the model in Matlab. The simulations shown in picture1.
- IDETC 2016:
have finished the abstract and introduction section. (picture 1 & picture 2)
December 2015
December 15
Wuming:
- Drawn and Exported the .stl file for the 3D printing part of the various surfaces for tumbling motion tests. Have coordinated Benji for the printing.
- Have run an exemplary model on the micro coil and magnet, the field and force on the magnet can be evaluated. (See pictures as the links below)
- Have been working with Sagar on the tracking/close loop, the old infrastructure is not functional.
December 8
Wuming:
- ufsmm prototypes prepared for the path planning tests, the magnets are attached both in Y and X directions.
- Revising the figures of previous section in the paper and wording remaining sections of the paper.
- Laser machining workshop quick notes for memo:
The machine is run by Prof. Babak Zaire's Lab at Birck 2081, current point of contact: Manuel Ocha, PhD student.
The machine info: PLS6MW multi-wavelength laser platform (http://www.ulsinc.com/products/pls6mw/).
Materials they have tried: Acrylic, paper, copper, zinc, PDMS(some smoking).
I haven't seen the advantage in terms of resolution, personal opinion.
December 1
Wuming:
- Setting photograph of new coil test bed with inverted camera: pic1, pic2, pic3.
November 2015
November 24
Wuming:
Update on the uFSMM journal paper:
- Have finished all sections, except: the path planning part, discussion, conclusion. Will show & touch base during meeting.
- Has done a few tests for path planning (move, avoiding obstacles).
- The stiffness on intuitive direction (Y direction) is still not ideally soft, but it should make sense as long as the the tip deflects. Will show the video and touch base during meeting.
November 10
Wuming:
- Has figured out an easy and new way for manual assemble the uFSMM. (pic1, pic2)
- Has placed and tested one robot with the pusher shape end-effector, raw videos on server, will touch base.
The deflection is small on the Y direction.
- The paper is under drafting. (glance)
October 2015
October 27
Sagar:
- Managed to get successful videos with one robot. Robot can now change direction depending on the waypoint. Videos are uploaded on backup drive: Z:\Current Projects\Microrobotics\Videos\microcoils automation videos\three coils_automatic
- Currently working on the multirobot videos.
Wuming:
- uFSMM: Have release some new (3rd gen.) prototype. (Picture). The uneven rate of release back side etching still exists obviously. But we'll just proceed anyway.
If necessary/desired, I feel confident to have SOI wafer for a try since lots of issues have been detected so far.
- uTUM: 1) Released some NdFeB one of the new pattern. 2) Attached some old pattern ones with the FotoFab Ni pieces.
October 20
Wuming:
- Fab of TUM: Finished the new run with Tumbling pattern, will release and attach the Fotofab Ni piece.
- Fab of uFSMM: New wafer working well so far, some harder for handling since the thin thickness, cracks after all the front pattern done, luckily larger part is OK. (Picture) Anyway, fab process will proceed with the larger part for backside pattern and release.
- Cleanroom is experiencing shutdown for all the equipments since Monday 10/19. The staff feedback told the facility will come back on by Friday, 10/23. So I'm hoping to have the first batch of new prototype by some time next week.
October 13
Wuming:
Fab run of tumbling one, with the long axe magnetic. The resultant magnetic property not so well. See following pics.
October 6
Wuming: updates & info for memo
- Wavelength of the UV light source at Birck
Both of the MA6 and MJB3 are set at 405 nm, and the 365 nm is also NOT filtered OUT.
However, the 250 nm light source is not present, for cross-linking the self-rolling polymer.
- Fab of uFSMM:
(1) Will use the new thinner (~300 um) & double side polished wafer. & feel possible to have the new prototypes less than 2 weeks (The ones without powder in PDMS).
(2) Checking out the DRIE machine for Helium Backside Cooling (HBC), the leaking is increasing which is significantly bad for the etching, did some dummy runs at the end of September.
The value now is now worse than the recommended, but still in the limit. Potential problem for the long etching.
- Training for PVD: Still pushing the magnetic PVD machine , couple of emails, but still even unable to get a training arrangement ...
- Tumbling one: The carrier box has arrived on Monday, will test the magnetic property roughly by hand and collect some in the carrier this week.
September 2015
September 22, 2015
Sagar:
-Working on finishing up the experiments with PCB. Image processing is the big issue.
-Rewriting the experimental section for the paper.
Wuming:
- Fab of uFSMM: One wafer front side pattern done, next to back side release. (The one is the pure PDMS without particle one). The particle one has the boundary wrinkle issue, will show pics and discuss during the meeting.
- Fab of new Tumbling: part of the small pillar structures are gone, only large size ones stand, next to release and collection. Will show the pics and touch base during the meeting.
September 15, 2015
Sagar:
- Finished some extensive open loop preliminary tests with the coils. Three coils do not work. We can use them as obstacles. The success rates are decent and depend on the initial position of the robot. With closed loop it will be more reliable.
- Finished the reviewers response. I have some confusion about one response. I could not find force values that the reviewer were suggesting. Forces depend on lot of conditions for a particular manipulation system.
Wuming:
- New ordered mask and petri dish supply received.
- Fotofab Ni part detached and tested, good for magnetic property (Video).
Current not tuned, ~ 1.5 A. using the new mini joystick hooked up by Bo, working good, but response has some delay.
- Fab of new uFSMM:
finish the PDMS curing and planing, to the second step for the frame part.
need to touch base before proceed, since the magnetic volume in the PDMS seems very low. (picture).
In the original solution, the powder: PDMS base ratio (volume) is already larger than 1:1.
Based on the results, I think there is no way we can have similar magnitude of magnetic force, even compare to the I-bar robot we use done by PVD. So even the stiffness lowered, i.e. 50%. but acting force we will have will be way lower.
So, do we wanna try with another PDMS run, or some other way ?
September 8, 2015
Sagar
Finishing up the response to reviewers and revising the paper. Will be done by the end of Tuesday.
Started writing the experimental result section for the journal version of DETC paper.
Getting prepared for the experiments.
Wuming:
- New mask order processed and shipped.
- Ongoing fab of new ufsmm prototype run:
1) New feature dimension indicates new result, like the etching rate varies obviously on the slim part and the more open area section of the robot. (Wednesday)
2) (Thursday) Proceed to the PDMS step, with 2-3 um Ni particle, but issue with powder not allowed in the clean room. Trying to do this week and figure out my next concern: the planing of PDMS & whether it will affect next step patterning of AZ photoresist that is thin (no more than 2-3 um).
August 2015
August 25, 2015
Sagar
Finished the matlab interfacing.
Started writing up the motivation for the papers. In the process of collecting relevant references.
uFSMM
Finished drawing the multiple masks.
August 11, 2015
Peter
Finished Becker mask - Wuming will edit it/add to it with the fading masks
Finished both coil systems and recorded video - issues with video files, but they're on the server
Discussion on fotofab
July 2015
July 21, 2015
uFSMM
- 3rd generation: new design and mask drawing is done, will touch base during the meeting for ideas and revision.
Then we can go ahead for the new mask and new run.
Marshall
- Laser cut complete for large scale designs. Will test them today with SMA.
- Hopefully finish design analysis on CAD today too.
Peter
I've requested a quote from fotofab - their capabilities should be sufficient to make blocks of nickel, but I need to know a price
Working on the mask for Aaron Becker's work - going slow (new to AutoCAD and mask making), but I'm getting better at it
Idea that distracted me for a few hours - using a track and swarm of robots for microassembly - has it been done?
Old coil repairs - I remeasured the coils, and determined that I previously didn't wrap enough wire, so I fixed that, but still have to:
Solder the wires
Either jam the cores into the replacement coils or machine them down a hair (I made a slight mistake with 3D Printing shrinkage)
Both of these will be complete TODAY - this needs to be out of the way
Sagar
- Finished testing with electronics. Its working fine with Arduino UNO. I would like to order an Arduino UNO as a backup.
- I am working on the presentation slides. Will send you the first draft in two days.
July 14, 2015
Sagar
- PWM verification is done. Getting ready for the experiment.
- Paper is almost done. Full draft will be done by the end of today.
uFSMM
- No further improvements with current prototype, the actual stiffness of the current prototypes is in the range about > 0.1 N/m.
- Revising the designs for the evolving of micro-scale uFSMM. * The narrowest beam width is 10 um, * footprint down to about 500 um, * one inserted step/mask, for filled PDMS with magnetic particles.
will show it during the meeting.
Peter
Too sick to come into the meeting today - my apologies - if you have any questions, don't hesitate to email me
Wuming is planning on having the mask design done sometime tomorrow, which should result in a new set of masks by the end of next week. Thus, I plan on waiting for these new masks before spending time and resources on refabricating an inferior set of uFSMM.
I'm playing around with the mask files to learn how to make those myself as well
Wire wrapping on old coils is complete, but still need to do the soldering (which should be much faster since I've had practice on the new coil system). Should be done in a day or two.
I can rearrange the coils so that the two replacements are across from one another. I'm going to go ahead and do that unless anyone disagrees.
July 6, 2015
http://www.kjmagnetics.com/categories.asp?gclid=CjwKEAjw8e2sBRCYte6U3suRjFESJAB4gn_gcQo4WKl7xv1Yk0AZz9URQiCZnNlDaW_fGuArNvyFwhoC7X7w_wcB
http://www.supermagnetman.net/index.php
uFSMM
- Testing: some trials of pushing with deflection, but the deflection is not significant anyway. (video)
- Further tests will be focused on making new prototypes with less glue, which means trying not to let the glue touching the PDMS section. If the coming out is still about the same, I'll try to figure out the paper based on these tests. Meanwhile we can start the efforts for evolving the design further.
Peter
Received DRIE/ASE training last Tuesday - should I create a repeat of the last uFSMM trial or a new design?
Finished soldering of new coil system - working on electrical testing and software (current limits due to heat) - should be done with this tomorrow
Communication error with laser cutting - need to order new acrylic sheet (necessary for mounting onto microscope, but not needed for initial testing)
Need to wrap the wire for old coils (we have plenty of 20 AWG wire) - should finish this by Thursday
Sagar
- Almost done with the survey paper: one introduction figure and challenges need to be polished.
- The electronics is fixed. Looks like Arduino DUE has some problems in I2C communication. Tried to communicate between two DUE boards as a test but failed. Will try with Arduino UNO (All the examples I found online used Arduino UNO).
June 2015
June 30, 2015
Sagar
- Finishing the paper. Working on Introduction and discussions
- Got the power supply. Testing the power supply with the big coils.
- Integrating the image processing for mobility testing of force sensing robot.
uFSMM
- When pushing the object, the micro force sensor will not deform. However, it does deform when hitting the force sensor probe to some fixed wall with higher current/field. (video)
- The specs for the testing.
magnetic volume - 600 X 200 X 25 (um)
Power input: a) pushing the object: I = 1.0 ~ 1.5 A.
b) hitting the wall with deformation: I > 6.0 A, field at center > 20 - 25 mT
- Forces evaluation: Force ~ 1.0 - 5.0 uN, Deflection ~ 10 - 20 um, Stiffness ~ 0.1 - 0.3 N/m. That coincide with the calibration of the micro force sensor.
- The smaller micro force sensor is more sensitive to the practice. Smaller force sensor has higher stiffness, due to the shorter beam and spring.
- Way to improve, avoid the latch (Si with PDMS) in the beam part, so the beam can be thinner, < 20 um. (pic)
Peter
Haven't received ICRA reimbursement from RAS/Dan Popa yet (figured I'd make a note of that)
Receiving DRIE/ASE training today
Waiting on heat shrink wrap to solder new coil connections (Daniel ordered some that I'll use - hopefully it's in very soon)
Need to perform one more laser cut - won't take much time at all
Working on 3D Printing old coil system (working on getting the scaling of the replacement part correct)
Note to self: make sure we have enough wire for the coils - we likely will need to order more.
Other than working with the coils, I'll need to work on getting the arduino software going for the new planar coil system (multiple pwm boards - need to figure out how to specify which one we're talking to)
June 23, 2015
Peter
ICRA expenses mailed to Dan Popa on 6/15 and entered into concur, though will not be submitted until the outside funding from IEEE RAS is received (doing all of this took much more time than I anticipated)
Emailed Mr. Rinehart for DRIE/ASE training - he got back to me quick but I somehow didn't notice his email, so I'll get trained on it next Tuesday afternoon.
Working on the new coil system - will finish by next week, barring any unforseen catastrophes. Remaining steps:
Laser cut acrylic sheet to hold the 3D-Printed part with the coils and iron cores
Create .dwg file (by doing a sketch as a part in solidworks and saving as) with layers (red is cut first, remove all text and construction lines)
Email Steve Florence and ask if I need any additional access, then schedule a time with him to perform the cut
Strip and solder the wires (myself)
Materials needed: red and black wire, dean's connectors, heat shrink tubing
Fixing old coil system - Broken coil(s) will need to be rewrapped - link
Weird idea that I figured I'd write down - single moving coil (just an idea, don't think it's worth spending time on, especially with our focus on the planar coil array)
Rather than having an array of coils changing their magnitude, a single coil that rotates around the workspace to change the direction.
Pros:
More reliable and better-known direction control (diagonal control with the coil arrays sometimes produces unexpected results)
More consistent magnetic field (since there is only one coil instead of multiple with a few differences)
The single coil could potentially be larger and more powerful, since there is only one
Cons:
More moving parts and points of failure
Would require a little more electronics (ideally a fairly powerful stepper motor and something to control it, with basic electronics)
Vibrations from motion of coil (there are ways of minimizing these)
Corollary - could there be a way to have a magnetic tweezer that traps magnetic particles at a point? Then perhaps to control motion the apparatus could be physically moved rather than the magnetic fields changed.
Sagar
Working on image processing of force sensor. Managed to get a single filled processed image that will be useful to get the deformation:
Finishing up the paper. Working on the figures as we discussed.
Talked to Bert. He did not get the parts as of Friday.
May 2015
May 6, 2015
MMC
Archit implemented manual control into the control software
Archit and myself have been working on moving the software to the laptop - we're making progress, but it's tougher than expected - but this is the first priority before Archit is gone
We got the control system in - I'll test it after the meeting before Archit comes in. I'm very confident it will work, since it's electrically identical to the last one.
Once it's working on the laptop, we'll work on implementing the remaining waypoint functions, which should be straightforward.
Peter's last final is Thursday night - after that, a lot more should get done
The hope is to have the software done by the end of Monday and devote the rest of next week to debugging/testing. The week after that, after Peter gets back on Tuesday night, will be devoted to packing/cleaning up anything else that comes up.
Coil System
I haven't forgot about this, but I DID forget to wear pants (instead of shorts) today, so I didn't get anything done like I was hoping to.
April 2015
April 28, 2015
MMC
Fixed the waypoint error - now has support for ~200 waypoints, which is sufficient, but I'm working on changing the framework of the software to allow for all three paths to be run at different multiples in different orders. Essentially, the plan is for the software to:
take four clicks at edges of the circle and convert that to waypoints for the robot to follow
take four clicks at edges of the rectangle and convert that to waypoints for the robot to follow
take three clicks at edges of the rectangle and convert that to waypoints for the robot to follow
have a way to order the paths and set the number of repetitions for each one (the plan is to have this in the GUI, but it can also be done by changing the code itself before each run)
Working on manual control with qt - Sagar has helped a lot with it, but we would like Archit to implement Sagar's framework into the existing control software so it's all in one package
Improving at removing robots from wafers
We have microassembly arenas from past competitions that should work for this one, but I need to double-check the dimensions on them to ensure they are correct
The new electronics should be done this week
It is crucial that we have Archit's help moving the software from the desktop to the laptop (as soon as it comes in). There are version issues with the current software, and we'll need to make sure we can get all of the correct versions or make the changes necessary to make the code work with newer versions.
Miscellaneous
Moved the storage container above the new table
Set up the new computer - everything seems to be in working order
Joseph set up the new air line for the table - it is quite level right now, but could be calibrated to make it more level
Peter is having phone issues - hope to get it fixed today, but for now, email is the best way to contact me
April 21, 2015
- Wuming
Guys, I'm traveling in the afternoon, so I just type the updates here to touch base.
1) uFSMM control:
Peter, I was just able to modify some GUI layout of the control, I'm not sure if you've check out the 'long period' issue. Anyway, I'll try to get it going when I get back later this week.
2) Variance design of the tumbling idea:
--- With the 250 um dimension, I tried to lay the two bell parts orthogonally (pic of the design), so one bell will experience driving force to +Y and one will experience torque. The idea is basically use the cube as driving part and PDMS arm as manipulation arm, it may also provide force information feedback, the deflection and stiffness of the design has not been numerically calibrated
--- Further, the two bell parts will have some initial interactive force & torque to keep the gripper open. I evaluated the force & torque in COMSOL.
(pic of the simulation). I'm not fully convinced by the results since that the force results varies more than even 20% with different meshing. I'm revising the air model and meshing. The results are getting some better with the error.
--- Anyway, I think with sophisticated control, the magnetic bells design with soft PDMS connection can have some initial gesture without exterior field and some interesting motion, or even 3D working mechanism. We can brainstorm for some ideas and designs with this sketch.
3) wafers run out:
--- I just used the last wafer to make some swimming pool for the multiple robots working space. The substrate is hard to have when the required substrate thickness (gap between robot and coil) is only dozens of microns. We wanna check out some thin ~~~ transparent paper and use PDMS wall to bound the fluid.
--- Peter, I think we wanna order a new box of wafer in our stock for future use.
The wafer can just be ordered through the online system here.
https://order.universitywafer.com/default.aspx?cat=Silicon
The product ID is #785 or #590, Qty. 25 which is a box. They are prime grade or test grade, not much difference, neither the price.
They are <100> orientation which is easy to cut.
I'm sure if any other fab usage in future, like if we wanna do back side etch, double side polished (DSP) wafer might be a better choice.
Anyway, you may have a quick touch base to decide.
MMC
The deposition was a success - I've tested the largest circular robot and it works well - I require more practice successfully removing robots from the wafer before I will know how well the smaller ones work, but I'm optimistic about their magnetic properties. The challenge will lie in how easily they are seen by the software, which I have yet to test.
I've been having trouble getting the robots off of the wafer - I think I need Wuming's assistance to get those going (I don't know how I got the one off last week). In the meantime, I tested one of the 100 small magnets we purchased, and it moves REALLY well. I posted a video on the server, which shows very fast movements in real time. I believe that the small magnet might be our best robot for both tasks, because it moves so well, although it has a diagonal length of ~350 microns, which is on the larger side for the autonomous accuracy task.
MMC To-Do List
Software
Support Multiple Paths (this is my biggest concern, but it's certainly possible)
Continuous Motion (should take very little work)
Move to Laptop (do we have the laptop yet?)
"Hardware"
Ensure coils are in good condition and ready to ship
Ensure control system is ready to ship
Determine need for spare parts
Construct Microassembly Event Arenas - need wafers
Testing
Must be proficient at setting up trials
Autonomous Event must work reliably before leaving
Microassembly Event will require enough trials to become comfortable with the strategy
April 7, 2015
MMC
Finally got training in Birck - will do deposition tomorrow
Finished Manual Control Program - standalone application to be run with FlyCap software.
March 2015
March 31, 2015
CM Robot
Wuming- (Not able to speak too loud, so I just type the work in a little bit more details below)
uTUM, tumbling microrobot validation:
1) Issues when running the force evaluation for microrobot: meshing size.
Ideally, the finite element size should be as small as, ie., 1/5 of the microrobot size. However, it's not OK for the computer (i7, 2.7GHz, 8G RAM) : too large problem, too many elements. Because we also build the magnetic field/air domain. It's meshing a 2" space with 25 um element.
Thereafter, I mesh the microrobot and magnetic field/air domain with different element size. The simulation series showed that the force simulation result is sensitive to the meshing size. Results varies with the meshing size of the air domain.
2) Regardless of the simulation error stated above, the simulation in COMSOL does NOT show convincing result for uTUM, which means comparing to the simple bar shape permanent magnet (pic), the dumbbell design does not show advantage in terms of momentum (force + torque).
So I would like to have some variant design further than the dumbbell design, ie, two magnetic bells connected with PDMS parts, so it will have some original curve when lying on some dry substrate. Needs further design and analysis.
3) ANSYS finally configured and installed successfully on my windows laptop, it's able to run various modules including the rigid dynamics one.
Electronics - Burt
As of last week, he was ordering parts for both the second 8-channel system and 64-channel system, including a box for the 8-ch system. That seems to be progressing well - we'll check with him after this meeting.
Bixing and Archit need a Work Order Form filled out for the soldering work, if it hasn't already been taken care of
MMC
Heard from Dave in Birck (he asked for my schedule for the training), but I haven't heard back yet. Hoping to get that done within the next week, but I've been saying that for like a month now.
Peter's been working on the software, but an exam last week slowed things down. The plan is to have manual control finished by Friday at the latest. Once that's done, practice tests will be run for the microassembly challenge.
March 24
Wuming-
1) Try to assemble the cubes to uFSMM (pic), same difficulty as before, but should have more magnetic momentum.
2) Doing the simulation for uTUM.
Microcoil
1) Almost done with the PCB design. I have few more things to figure out.
2) Started interfacing (using camera SDK) with the Allied vision camera. It requires a 12 V power source. The framerate is about 60 fps. The field of view is also good for the PCB coils.
March 17, 2015
Petri Dish Coil
I finished wrapping of another coil, but ran out of 18 gage wire. There are two options:
Order more 18 gage wire for $38.70 from McMaster-Carr here
Use 20 gage wire, which would increase the number of turns, and thus the magnetic field as well as construction time, but would get hotter quicker with lower amounts of current.
I'd prefer to use the 18 gage wire to keep things consistent (otherwise the two 20 gage coils would require different amounts of current to run)
Also, I talked to the Herrick shop about cutting the cast iron rod into four pieces and machining it to fit the collar (tests with the cast iron core were successful), and it should be easy to do , but they need an account number to perform the work.
MMC
PVD Machine is down (contamination) - I'm first in line to receive training. I'm hoping it'll be fixed within a week, but I have no idea.
When I get the training, I'll use the Startup Discretionary Account for any charges.
In the meantime, I'll work on making the left coil tighter and play with the software to make the robot motion continuous rather than start-stop.
March 10, 2015
uFSMM: Several prototypes have been glued, although not that well but attached. Will try to do the testing when the manual control is OK.
MMC
Have been in contact with Dave Lubelski in Birck for PVD training - he will get back to me tomorrow or Thursday to schedule the training - should be done by next week at the latest
March 3, 2015
Fabrication during last week.
(1) pegs and smaller pattern of substitute for I-bar robot. (picture at a glance)
(2) Collection of uFSMM
MMC
Wuming completed peg pattern for replacement robots
Waiting on training - same issues as Sagar from a few months ago - communication with Birck staff is slow
Culture Dish Coil
Tested one coil - it moves a robot in water very slowly at 2.25 A. We'll need to use a core - next step is to test the cast iron core (which is missing). If that doesn't help enough, then we'll want a Cobalt Iron core or I can resize the coils to allow for more turns.
February 2015
February 24, 2015
https://www.youtube.com/watch?v=730D9-QtUi0
uFSMM : Mechanical strength improved in an obvious way, more survival from tweezer, but still the same for the glue. Maybe better with device like micro-gripper, but much more likely with process during MEMS fabrication.
Thermo Bond: schematic showing this technique. (picture)
Tumbling microrobot: try to redo the theoretical part, in COMSOL, (picture), haven't found the way to do the kinematic demo, only force can be evaluated.
Microcoil
The second round of Ti deposition is scheduled for Next Wednesday. The last week appointment was cancelled since Kenny was sick.
Performed some preliminary experiments with PCB coils. It is able to move the robot. Smaller robots exhibit reliable motions.
February 17, 2015
Microcoil
Tried the copper electroplating with positive photoresist patterning. However, the photoresist seems to be peeled of at the locations of small cross-sections making the areas open for electroplating. Following are some pics:
I will make another try with soaking the positive photoresist with Chlorobenzene before patterning to make the layer resistant to the attack of Cu solution.
Bert has started making the 64 coil system electronics.
MMC
Field is constructed and looks good under the microscope
Robot is manually traversing the field well - video captured of the robot moving manually in the rectangular shape
Archit is still making changes to the autonomous code - successful autonomous video not captured yet
February 10, 2015
"CM" Robot
Redesign the chest - ( the screw holes are too fragile)
Testing the SPY camera under 5V condition (Normal operating voltge: 8V)
Learning how to use liquid rubber to mold the leg
MMC
Control Software is successfully communicating with Arduino, but as of Saturday, there are still issues with the automation - Archit is working on it
Issues with robot sticking to old field (probably because Sagar and I made it dirty by using oil to test the one-channel control system prototype) - hoping this gets fixed with a clean new field - can Wuming help perhaps after this meeting?
Current plan until Feb. 22nd:
Fix automation so that the robot can move to waypoints in horizontal/vertical steps, then backup the software that does this
Take a video of the robot navigating the Rectangular trajectory (I think this would satisfy the requirements for the video - am I correct?)
Use the microscope to take pictures of the microrobot for scale
Change automation algorithm to allow for diagonal motion
Test and refine algorithm towards the goal of moving the robot at the angle for the triangle, then backup the software that accomplishes this goal
Take a video of the robot navigating the Triangular trajectory
Test and refine algorithm towards the goal of moving the robot in an octagonal shape (for the circle), then backup the software that accomplishes this goal
Take a video of the robot navigating the circular trajectory
mobility arena done: pic
need to do make the pegs.
Microcoil:
Copper laminate trial under way: pic1, pic2
waiting for another sample and then try etching.
February 3, 2015
uFSMM
New prototypes collected, around 30, yield is still low. back side release etching is even harder than before.
MMC
Microassembly Robots - my main concerns are:
Directionality - It is difficult to control the orientation of the robot in addition to its location
Sticking - the robot should not permanently attach to the triangle upon contact
I may be overestimating this as a concern
Sufficient contact with triangle - the robot needs enough contact to reliably manipulate triangles.
Construction - the robot needs to be able to be removed from the wafer without being damaged or lost
Size/Maneuverability - it's worth considering a robot small enough to maneuver around a centered triangle in the small area of the channel - about 200 microns long. This may be impractical
Deadline extended a week, but still targeting 2/15 for video completion
RAS travel assistance application - do we need to do anything for this?
Control System has arrived - Arduino software almost done
Have we worked on computer control software? This needs to be working ASAP - I expect there to be a lot of required testing
Masks submitted - we need the arena for the video, so we'll need to fabricate this as soon as the masks come in
Need to capture images of robot size for video (just a reminder to myself)
January 2015
January 27, 2015
uFSMM
current run in the release etching step.
MMC
For the field: i'll edit the drawing today and send it out by tomorrow for the transparency mask. (Wuming)
The assembly field dimension is the same as before, the peg size is not the same as what we have before.
January 20, 2015
uFSMM
Prototypes in previous overshoot run are too thin for tweezer handling during collection. (See pics)
Another run has already been underway (PDMS filing and planing finished).
MMC 2015
VIDEO DUE FEBRUARY 15TH
Autonomous Trajectory Task: The microrobot musttraverse one or more of the representative autonomous trajectories listed in the Autonomous Moibility & Accuracy Challenge task description. The trajectories must be traversed on an arena meeting the competition specifications. Videos must be time stamped showing that the traversal is completed in less than 2 minutes. The robot must be controlled autonomously. The qualification video must be accompanied by a demonstration that the microrobot meets the dimensional requirements listed in the official rules.
Waiting on micro-3D Printing guy as of Friday - Peter will get PVD training when the robot is nearly ready to coat to ensure the training is remembered (trainer's rule)
Alternative design?
Planning on using existing coils with new control system
Existing camera seems to be having display issues, it might need replacing
1-channel control system has been tested and Bert is working on 8-channel version - we're hoping he'll be done within 3 weeks (by Fri, Feb. 6)
Can we use the existing control system for the existing coils?
Tasks to be completed:
Microcontroller-level control software (between computer and coils) needs to be written - Peter will do this
Computer control software (utilizing camera vision) needs to be written - Peter will need help with this
Field mask needs to be finalized, populated, and fabricated - Peter will do this ASAP
Material for Microassembly Triangles?
Cm scaled "Ant" - Robot
Interfacing with DC motor, motor controller and MPU has been implemented
A second version of the chassis(body) design has been designed on CAD
Communication with GPS device and HD Camera Cape is being studied
There are two option available for wirless communication with a laptop/host :
i) Using a pair of Xbee Transceivers for individual serial data communication
ii) Mini-USB WiFi Module for connection with the LAN ( Would require use of standard network protocols for host-host communication)
January 7, 2015
Coils
3D Printed - Working on wire wrapping
December 2014
December 17, 2014
Mobile Microrobot Project
The Cu electroplating works fine. However, removing SU8 is difficult. I have tried PG remover to remove SU8. It can remove SU8 but it also peels the Cu off from the surface. We should make two more masks to use AZ (positive photoresist) instead of SU8 (negative photoresist).
Coils
Design Finished, screenshots below. Will 3D print today (Wednesday) assuming our new Herrick printer cooperates with me (I've never used one myself before. If there are any issues with the printer that you know of or if you aren't comfortable letting me figure it out, let me know)
Ordered Cast Iron Core - still looking into Cobalt Iron and Soft (annealed) Iron cores
Ordered collars
If everything goes right and nothing breaks, I hope to finish construction of the coils before I leave Saturday evening.
Mobile Microrobotics Challenge
Proposal Accepted
Video due March 1st
Will hit this hard after I get back the evening of Monday January 5th
December 10, 2014
uFSMM
- Fab: front side photolithography, PDMS, etching are all completed, only the backside etching left to go, waiting for the 6" oxide wafer as DRIE substrate.
- Paper: to redo the theoretical values of the micro force sensor, since more PDMS portion for the new design.
Coils
Done with CAD files - need to 3D Print
Select Coil Wire
Order and cut Iron Cores
November 2014
November 25, 2014
Mobile Microrobot Project
PVD training is completed. The fabrication cannot be started since David Lubelski did not give time.
Basic serial communication with Arduino Due is done. There needs to be a delay in the communication.
Working on integrating D* Lite planner with the simulation.
The fabrication plan is updated on the server.
uFSMM
Fab of 2nd batch uFSMM is under way. silicon frame has been patterned over PDMS, waiting for another DRIE (1 day), and back side release and collection (about 1-2 week).
Some quick glance below.
Present questions: based on the previous runs of the DRIE, the etching selectivity of the machine (SiO2:Si) is not that well, so the oxide wafer substrate is consumable.
Can we order a box of 6" wafer with thermal oxide after Thanksgiving.
November 19, 2014
Mobile Microrobot Project
Done with the first SU8 test layer on the wafer. The patterning looks good.
Started the matlab interfacing with Arduino Due.
Setup a four coil simulation environment. The forces are verified with the equations in MIT notes.
Some hand calculations of the inductance and resistance show that the response time (tau = L/R) of the coils (LR circuit) is in the order of microseconds. That means, the frequency of the system is limited by the maximum data transfer rate of the controller. However, for bigger coils the response time might be higher.
TO DO:
The PVD training is pushed to Thursday.
The COMSOL analysis will be done this week.
Wuming-
2nd generation uFSMM masks are ready, starting fab.
The pattern sample on glass slide has been made (AZ1518), will bring to Peter check out to decide how to prepare the arena.
Cm-scale 'insect' robot
New frame design with motor and battery mounting
New Leg design
Leg 3D print Test
November 12, 2014
Mobile Microrobot Project
Masks are done.
Top masks for the SU8 finishing layer and the wire bonding are done.
Lithography training is completed
Simulations are corrected with new equation that takes the volume of the robot into account.
To do:
We will start lithography on Thursday
Finish the training for PVD
Verify the simulation qualitatively with Solidworks COMSOL multi-physics Engine
uFSMM:
2nd generation masks are revised and verified, currenly in the manufacture queue.
Mobile Microrobot Challenge 2015
Peter is working on the drawing of new field pattern.
I have got the top masks for the micro coil, and will edit them with this field mask and outsource them then. (Wuming)
Suggestion: some 5"X5" glass pieces are useful for the transparent mask, that will be handy since not detach and tape every time, also less chance of stain.
November 5, 2014
Mobile Microrobotics Challenge 2015
Working on mask for arena, the former mask files have been uploaded onto the ECN server.
Team proposals due December 1, 2014.
uFSMM
Design for the manipulator shape on the uFSMM may be varied.
October 2014
October 29, 2014
New lab server:
MAC:
smb://spiralab.ecn.purdue.edu/MSRAL
PC:
\\spiralab.ecn.purdue.edu\MSRAL
“Map network drive”, use the above link
Login with your Purdue Career account credentials, double check the domain is "ECN", not your own computer.
Current Projects --> Microrobotics
Mobile microrobotic project:
The mask files are submitted and training on lithography is requested.
We are deciding on microcontrollers for the electronics design. We require a 16 bit microcontroller. Arduino Due is a good choice. BeagleBone might be too much.
Talked with Gale Lockwood about wirebonding. She is confident that we can wirebond with copper as long as we have a layer of gold of couple of nanometer on top of copper. The wire has to be of gold of 25 micron.
uFSMM:
Masks for the 2nd generation try, more anchor dimensions for the easy calibration.
Cm-scale 'insect' robot
Group Google Doc Link: https://docs.google.com/document/d/1-X6vGLG6yJLwfgI2lzDtAVID_IhI3j6LyjDGpyg5KXo/edit?usp=sharing
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Final parts selection/ suggestion
Motor Driver :http://www.robotshop.com/en/pololu-dual-dc-motor-driver-1a-4-5v-3-5v-tb6612fng.html
MPU: https://www.sparkfun.com/products/11486
Motor：https://catalog.precisionmicrodrives.com/order-parts/product/212-108-12mm-dc-gearmotor-22mm-type
BBB reference manual: https://github.com/CircuitCo/BeagleBone-Black/blob/master/BBB_SRM.pdf?raw=true
HD Camera Cape: http://54.186.12.74/Order/index.php?route=product/category&path=33_59
With option of replaceable lens and cable.
Or Wireless Camera: http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2047675.m570.l1313.TR11.TRC1.A0.H0.XMini+Wireless+Camera&_nkw=Mini+Wireless+Camera&_sacat=0
Option 2:http://www.adafruit.com/products/746
October 22, 2014
Cm-scale 'insect' robot
Decision between DC brushed and brushless motors (or stepper motor?)
DC Pager motor option:
Choice of MCU - H-Bridge Motor Controller (T-bone Cape- should include Motor Encoder)
Cape options: http://elinux.org/Beagleboard:BeagleBone_Capes
If can find correct capes (IMU, motor controller), may not need separate flight controller unit (Pixhawk, APM)
Other electronic parts required - Inertial Measurement Unit (IMU), Motor Encoder
Communication with SPI or SCI(UART bridge)?
Specifications required for Torque, wireless Wifi/Bluetooth connection with Mobile/Laptop
Mobile Microrobot Platform
Wafer order submitted - 10/21
Wafer carrier order submitted - 10/21
Need quote to order SU8-5
Need to confirm what copper/gold etching solutions we need to order - or can we use them in Birck?
Do we need electroplating solutions? other supplies (copper tape)?
Once coil design is set, need to design PCB for wire-bonding and outside control
Cross-check mask design with wire-bonding expert
October 15, 2014
Mobile Microrobot Platform
Change the coil design:
Increase the spacing between the tracings
Arrange the pads in two different columns (that will make the wire bonding easier)
Change all the circular features into polygons
Make the critical dimensions larger than 5 microns
Fonts for the pad numbering need to be changed with more polygonal shapes
UPDATE:
Had meeting with Aamar at Birck on Oct 17. The design looks good. He has forwarded it to Mike (who will do the mask) to have a final check. If he is okay with the design we will populate them on the wafer with alignment feature and dicing mark.
Collect information about electroplating (Do we need separate masks for electroplating?)
Had meeting with Michael Sinani about electroplating on Oct 17. He said, we need seed layer for the vias' as well. He is more concerned about the soft curing temperature of the SU8. This must be greater than that of AZ 9260 (which is 100 C). The reason is AZ 9260 will go on top of SU8. If SU8 curing temperature is less than 100 C there will be thermal stress developed in the bottom layer that will damage the part. He will be happy train us on the use of copper bath.
Need ordering info for all necessary fabrication materials/supplies:
Wafers with oxide layer : https://order.universitywafer.com/default.aspx?cat=Thermal%20Oxide&diam=100mm
Notes: We usually get a box of 25 wafers.
I think the one should be product # 1855.
There are also similar ones with different resistance,
SU8-2: http://microchem.com/pdf/SU8_2-25.pdf
Notes: Maybe the SU8-5 is the best one, I never noticed they have this one except 2 and 25.
My experience is you may wanna reach the thickness in one spin and softbake step. And you also wanna have the thickness with a speed not that slow, like hundreds rpm, that will make the photoresist hard to reach the wafer fringe and make the spin layer not uniform, especially the viscous one.
Electroplating solution?
Copper/gold etching solution: I don't have the experience to etch copper. I just did a quick search.
Below should be the one you wanna have, either one I guess, they have the same solute, you may check it out or have a careful search.
http://www.sigmaaldrich.com/catalog/product/aldrich/667528?lang=en®ion=US
wafer carriers: http://www.tedpella.com/storage-boxes-bags_html/wafer-carrier-trays.htm
Notes: The product # 1395-40
tweezers: You can get the tweezer and glass ware from Ira Young. I can't remember the room number, his office is at Birck, 2nd fl, the west end ? next to the stairs and elevator.
For the photoresist, you also have to address him, he bring the chemical bottle into the cleanroom and with label of your name.
Electronics: We will go with 16 bit design (that will give us fine resolution)
Cm-scale "insect" robots
IMU option (review says it works well with Beaglebone Black):
http://www.robotshop.com/en/6-dof-gyro-accelerometer-imu-mpu6050.html
Camera option:
October 8, 2014
Electronics Systems
Peter and Sagar working with Bert on electronics system designs:
64 coil control system
Need to control the current magnitude and direction, and read actual current levels for eventual closed-loop control
Typical max operating current of 500 mA / coil --> over-design to handle 1A
Sagar to check these numbers
6-8 coil, high current system
Need to control the current magnitude and direction, and read actual current levels for eventual closed-loop control
Typical max operating current of 6A / coil --> over-design to handle 8A-10A
Currently only have 6 coils to control, would like option to control as many as 8
For both cases, would like to have all processing done on-board PC (desktop or laptop)
Just communicate to electronics (with on-board microprocessor) information regarding what coils to turn on, current magnitude, and current directions
Communication can be serial, USB, ethernet, etc.
Separate power supply units for each box
uFSMM
Update electronics (see above)
New coil system design --> Peter has files from Wuming
Real-time tracking: displacements, position, orientation --> Archit
uTUM
No updates
Mobile Microrobot Platform
Sagar to finalize mask CAD files
Need to modify to accommodate for pads on all three layers of design
Modify spacing and arrangement of solder pads
Investigate requirements/typical layout for wire-bonding
Work with Bert to design PCB to go around MEMS substrate to connect to wire bonded pads
Set up meeting with Birck to discuss masks and fabrication details
MMC
Website: https://sites.google.com/site/mobilemicrorboticschallenge/
Proposal: 12/1; Video Submission: 2/15
Peter to create mask files for new arenas
Micro/Nano 3D printing
First batch of parts ready next week
Need to figure out how to make them magnetic:
Nickel sputtering/deposition?
Dope photoresist at beginning?
New designs for MMC robots
Cm-scale "insect" robots
Use KodLab Tobi robot for inspiration: http://tobirobot.wix.com/tobi
Use more powerful CPU that can have wireless control capabilities, camera and other sensor input: BeagleBoard Black
Add motor drivers (cape/shield)
3D printed housing / legs
Need on-board IMU --> look at open-source autopilot boards for this (ArduPilot, PixHawk, etc.)
camera link:
October 1, 2014
uFSMM
Deadlines: IROS (3/6) or IDETC (Jan 12/26) or Journal Submission
Work on new masks for next generation - Wuming/Peter
Real-time tracking of displacements for force control - Archit
Real-time position and orientation tracking - Archit
New coil system design - Peter
New electronics design/order parts - meet with Mike Logan
uTUM (Journal paper revision - Jan 1)
Maybe add new camera for 3D?
Tracking - position and orientation - Archit
MMC
Official rule are up! - Peter
Proposal: 12/1; Video Submission: 2/15
Mobile Microrobot Platform
Need CAD files to make mask - Sagar
Need order in for SU-8 - 2 (negative resist)
Use AZ9260 instead of KMPR (positive resist)
Update fabrication plan based on meeting (2 options)
New meeting with Birck once have 2D CAD files ready
Micro/Nano 3D printing
Check status of parts
New designs for MMC robots
Cm-scale "insect" robots - IDETC M&R (Jan 12 / 26?)
See Ron Fearing at UC Berkely, Rob Wood @ Harvard, Tiny Terp for design ideas
Low cost, off-the-shelf electronics (open source - Arduino?)
Easy to manufacture - 3D printed parts, laser cut parts
Direct drive motors? Need complicated mechanisms? (see Robot Shop for small motors (pager motors) )
September 2014
September 24, 2014
uFSMM
Fabrication process improvements:
Scale down
Calibrate on-wafer with force-sensor
Micro-magnet press-fit into PDMS for magnetic body
Dope PDMS with magnetic particles
Real-time vision tracking of deformation - need cleaner images/videos
Experiments with polystyrene spheres (get larger sizes)
Experiments with cells/tissues - find collaborator at Purdue
Way-point path following experiments
Implement force controlled manipulation
New coils system design to go around petri-dish
New control electronics
uTUM
uTUM real-time tracking
Autonomous control of tumbling cycle
Way-point path following experiments
MMC
Autonomous path following with position control for accuracy (I-bar)
Microassembly task
Microrobot Swarm Platform
Meeting @ Birck on Friday, 9/27/14 to discuss coil fabrication plan (meet in MSRAL@Spira at 10:45am)
Set up meeting with E-shop regarding electronics
Talk to Dr. Chiu's group regarding inkjet printing of coils (copper, multi-layer?)
Micro/Nano 3D printing
Send CAD files to Xu's group for printing
Helical Swimmer robots
Microassembly components