If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Open-source projects continue to grow in popularity alongside open-source educational resources, software, and hardware tools. The impact of this increased availability of open-source technologies is that end users are empowered to have greater control over the tools that they work with. This trend extends in the life science laboratory space, where new open-source projects are routinely being published that allow users to build and modify scientific equipment specifically tailored to their needs, often at a reduced cost from equivalent commercial offerings.
Recently, we identified a need for a compact orbital shaker that would be usable in temperature and humidity-controlled incubators to support the development and execution of a high-throughput suspension cell-based assay. Based on the requirements provided by staff biologists, an open-source project known as the DIYbio orbital shaker was identified on Thingiverse, then quickly prototyped and tested. The initial orbital shaker prototype based on the DIYbio design underwent an iterative prototyping and design process that proved to be straightforward due to the open-source nature of the project. The result of these efforts has been the successful initial deployment of ten shakers as of August 2021. This afforded us the scalability and efficacy needed to complete a large-scale screening campaign in less time and at less cost than if we purchased larger, less adaptable orbital shakers.
Lessons learned from prototyping, modifying, validating, deploying and maintaining laboratory devices based on an open-source design in support of a full-scale drug discovery high-throughput screening effort are described within this manuscript.
Over the past two decades, end users in a variety of industries have seen a change in the problem-solving process when it comes to addressing technical challenges that require specialized instrumentation. Where users once asked, “Where can I go to buy that or who can build this for me?”, users now find themselves asking “Has someone already solved this problem or can I design something myself?” [
]. This change of approach in problem solving has been brought on by increased access to specialized, domain-specific information and tools. Platforms such as YouTube have greatly reduced learning curves and the increased availability of open-source software and hardware tools have made various technologies much more accessible [
]. Examples of such software and hardware tools include hobby-grade 3D printers, laser cutters, and computer numerical control (CNC) machines. All of which can be easily coupled with output from computer-aided design (CAD) software, electronic design automation (EDA) software, and similar tools. High levels of interest in open-source scientific hardware can also be seen via the establishment of the new HardwareX journal, which focuses specifically on open-source scientific hardware [
Using open-source project repositories to find laboratory equipment
In late 2019, a high-throughput assay was in development in the Lead Identification laboratory in the department of Molecular Medicine at Scripps Research [
] that utilized 250 mL Erlenmeyer flasks for cell culture. The protocol for this assay required flasks to be consistently shaken inside a Thermo Scientific (Waltham, MA) Forma Steri-Cult CO2 incubator over a period of 3 to 4 days. The cells used for the assay were cultured in suspension under constant shaking to decrease the normal doubling time by half, which is approximately twelve hours compared to the 24 hours for contact-inhibited adherent cells. Another reason to culture cells in suspension is to have a better distribution of oxygen in the media to avoid dead zones. To meet the volume of cells needed to support daily HTS (high-throughput screening) efforts when the project was active, seven flasks would need to be simultaneously shaken inside of a single dedicated incubator. Given the space constraints of the incubator and the number of flasks in need of concurrent shaking to support an active HTS campaign, we were specifically looking for an orbital shaker that had a small physical footprint and that would allow multiple shaker units to be installed in a single incubator.
To identify an orbital shaker that met the specified criteria, we made an effort to search for existing projects published in online 3D printing repositories. These repositories, such as the NIH 3D Print Exchange [
], provide search interfaces which allow users to specify what types of projects or devices they are interested in. Users can use broad search terms such as 'laboratory equipment' to see all projects that have been tagged with these terms. Alternatively, searches can be performed with more specificity if the type of device or project of interest is known. For this project, searching for the term 'orbital shaker' on Thingiverse led directly to a page for a project titled 'DIYbio Orbital Shaker V1.0′, which was published on November 8, 2017. This specific orbital shaker was of interest as opposed to others published at the time, or to commercially available orbital shakers [
], due to its small footprint, low cost of parts, and design simplicity. The DIYbio Orbital Shaker featured a straightforward and simple design: a shaking platform and stepper motor connected to a stepper motor controller and microcontroller with a potentiometer to allow user adjustment to the motor speed. The enclosure and all mechanical components for the project, except for the bearings and fasteners, are 3D printable. With all parts in hand, as seen in Fig. 1, it is possible to assemble an orbital shaker in less than one day.
One example of the practical benefits common to open-source projects include documentation and feedback provided by community members. For the DIYbio orbital shaker, community members created additional project documentation, including YouTube videos detailing part-by-part assembly [
]. These videos, along with the project details available on Thingiverse, provided an opportunity to see feedback from users who have already gone through the assembly and testing process.
The DIYbio orbital shaker as published on Thingiverse included all the 3D models necessary to print functional components for an orbital shaker. The Thingiverse project page also provided a complete list of commercially available electromechanical components which could not be 3D printed such as the stepper motor, motor controller electronics and bearings. The project creator, ProgressTH (http://www.progressth.org/), also included wiring diagrams and assembly notes that made assembly of the first prototype device very straightforward. The provided 3D models included a flat shaking platform to which labware could be affixed with foam-backed adhesives.
Modifying an open-source project
While fully functional, the initial DIYbio Orbital Shaker prototype as fabricated from the Thingiverse instructions had room for improvement in a few areas: 1) Adding a method to mechanically capture labware on the orbital shaker platform, 2) simplifying the electronics to replace the hand-wired electronics protoboard with a custom printed-circuit board to make production of future units easier and to simplify maintenance, 3) modifying the motor coupler to enable shaking of larger items without causing mechanical failure of the coupler as observed during testing of initial prototype, 4) redesigning components to replace use of friction welding and adhesives with threaded, heat-set inserts and press-fit bearings and 5) improving upon the original CAD models.
The full list of parts needed to assemble this updated orbital shaker can be seen in Table 1.
To make the desired modifications to the originally published DIYbio Orbital Shaker V1.0 design, the CAD files were ported to from Google SketchUp (Mountain View, CA) to Autodesk Fusion 360 (San Rafael, CA) by recreating each component in Fusion 360. We found that having the components in the SketchUp format made them difficult to modify as desired and determined that it would be easier to recreate them in Fusion 360. Having access to the CAD source files is important not only for the purpose of making modifications to the published design, but also to correct potential errors in exported STL files or to optimize components to minimize the amount of 3D printed support material needed, both of which are problems commonly found with files on 3D printer project repositories.
Modifications made to the 3D models include: 1) changing the potentiometer mounting point from a small 3D printed potentiometer mount component inside of the case to being mounted directly to the inside surface of the enclosure, secured with a nut and washer, 2) adding screw mounting points for a custom motor control PCB (printed circuit board), 3) modifying the design of the motor coupler to be larger to prevent part failure, 4) increasing the diameter of the hole in the top cover of the shaker to allow the motor coupler to pass through to simplify installation and removal for maintenance and unit assembly and 5) replacing any components that were held together using glue or friction welding with heat-set inserts or press-fit installations. All 3D printed components were fabricated using a Prusa i3 MK3s 3D printer with PLA plastic. 3D models were prepared for printing using PrusaSlicer with 0.3 mm layer height and 100% infill.
During initial testing, the 3D printed motor coupler was found to be a point of failure, frequently breaking shortly after installation. The following attempts were made to address this issue: 1) redesigning the coupler to be thicker, 2) printing the parts in PETG plastic 3) orienting the coupler CAD model in different positions prior to 3D printing to change where the layer-to-layer thermoplastic printing stresses would occur and 4) ordering bronzed-silver steel couplers from Shapeways (New York, NY) [
]. Redesigning the coupler to be thicker did not solve the problem alone, nor did changing the plastic used from PLA to PETG. We were successful in resolving the issue by combining the thicker coupler design with part orientation optimization in the 3D printing slicing software [
]. Bronzed-silver steel couplers also proved to be successful in preventing coupler part failure but were expensive ($38.35 each) relative to the rest of the project and difficult to tap and thread. CAD renderings of this revised prototype and the exploded CAD model of the revised prototype can be seen in Fig. 2.
Electronics
The soldered protoboard of the initial prototype, seen at the top of Fig. 3, consisted of an Arduino compatible ProMicro 5V microcontroller from SparkFun Electronics (Boulder, CO) [
]. These components were wired to a stepper motor, potentiometer and power supply to complete the functional prototype. To simplify the wiring and give the electronic components a better fit inside of the orbital shaker case, a custom PCB, seen at the center of Fig. 3, was designed using the open source KiCad EDA [
] software. This software allows users to capture electronic schematics and create designs for printed circuit board production. The design of the circuit board allowed the ProMicro microcontroller and DRV8825 stepper driver to be plugged directly into the PCB and provided convenient wiring terminals for the motor, potentiometer, and power wires as seen at the bottom of Fig. 3. The PCB was also designed to have four screw holes to allow it to be mechanically installed inside of the orbital shaker enclosure. The completed PCB design files were ordered from a pooled PCB production service called PCBWay (Hangzhou, China) [
]. Utilizing a custom PCB allowed us to keep the footprint of the electronics as small as possible in order to keep the enclosure size of the shaker to a minimum, thereby maximizing the number of shakers which could be utilized in a single incubator.
Fig. 2Left - Modified CAD prototype. Right - exploded view of modified CAD prototype.
Fig. 3Top - Hand-soldered protoboard from first prototype. Middle - Front and back of custom PCB with microcontroller and motor controller installed. Bottom - Updated prototype with updated CAD enclosure and custom PCB installed.
All KiCad design files for the PCB, including the Gerber formatted files used for PCB ordering and production, along with updated CAD models for the enclosure and Arduino code, can be found on GitHub at the following URL: https://github.com/pierrebaillargeon/Scripps-DIYbio-orbital-shaker. Providing all source and production-ready files allows users who may have varying levels of experience with and access to various prototyping methods (3D Printing, PCB production, etc.) to interact with the project at a level appropriate to their level experience and interest. This means users can modify CAD files and export their own customized STL files for 3D printing or take the provided STL files and 3D print them directly without modification or upload them to an on-demand prototyping service like Shapeways. Similarly, users with limited electronics experience can take the provided Gerber files and upload them directly to a pooled PCB production service of their choice and have custom PCBs delivered without ever modifying the electronics schematic or PCB layout.
Creating documentation for open-source projects
An important aspect of creating or modifying an open-source project is documenting the work done to help other users reproduce the work and facilitate future modifications or maintenance. An example of this, mentioned earlier in this manuscript, was the video documentation published on YouTube that provided assembly and testing observations. To document the assembly steps for the modified DIYbio orbital shaker described in this manuscript, we have created a Dozuki [
] How-to Guide that can be accessed at the following URL: https://scrippsresearch.dozuki.com/. Dozuki Guides are web-based interfaces which allow instructions and photos to be easily marked up and organized for users to follow when performing a set series of tasks. These guides are popular in the hobbyist 3D printing community and commonly seen with build guides for Prusa [
] 3D printers among others. One benefit of self-help style guides like those seen in the Dozuki style is that they enable end users to maintain and troubleshoot their own equipment if desired, resulting in increased system uptime and reduced maintenance costs compared to traditional service models which require a service engineer to arrive on-site for troubleshooting and repair.
Performing the screening assay
The assay which utilized these orbital shakers to grow the cells was developed and miniaturized as a multiplex dual assay mode into 1536-well format. Cells were routinely cultured at a density of 0.4 × 106 cells/mL in a total volume of 40 mL of media in a 125 mL Erlenmeyer flask at 37°C, 8% CO2 and 80% relative humidity. Cells were placed in the orbital shakers for 3-4 days at a speed of 125 rpm with a shaker throw of 19.05 mm (0.784 in.) as seen in Fig. 4. Prior to the assay, cells were seeded in non-selection media and harvested after 4 days in culture. On the day of the assay, cells were washed twice with ThermoFisher Scientific Gibco (Waltham, MA) Scientific DPBS catalog # 14190144 (Dulbecco's phosphate buffered saline without Calcium and Magnesium) and re-suspended by hand to the appropriate cell density in assay media. The pH range of this reagent is between 7-7.3 according to the specifications from the manufacturer. This reagent is a balanced salt solution, and it was used to wash the cells prior to diluting the cells for the experiment. We used 45mL of DPBS to wash the cells for each wash step.
Fig. 4Modified DIYbio orbital shakers in use inside of incubator.
Prior to start the assay, cells were seeded in non-selection media and harvested after 4 days in culture. The non-selection media does not have the selection reagents that are used to grow the cells. On the day of the screening assay, cells were washed twice with DPBS and re-suspended by hand to the appropriate cell density in non-selection assay media. On the day of the experiment, 5 uL of the cells preparation were dispensed using an Aurora Biosciences (San Diego, CA) Flying Reagent Dispenser (FRD) into a white solid bottom 1536-well microtiter plate. The sample field and the low controls wells contained 80 cells per well and the high control wells contained 2,500 cells per well. After the cells were plated, compounds were transferred from a 1536-well compound source plates into 1536-well assay plates via GNF robotic platform (La Jolla, CA). The assay was incubated for 72 hours at 37°C incubator, 95% relative humidity and 8% CO2. This assay has two detection modes, the TRFRET, and the Luminescence. After the 72 hours of incubation, 10 nL of the supernatant of the cells treated with compounds was transferred via GNF (La Jolla, CA) [
] 1536-well pintool into a new white solid 1536-well microtiter plate containing 2.5 uL of the non-selection media. Then, 2.5 uL of the detection reagents for the Time-Resolved Fluorescence Energy Transfer (TRFRET) were added according to the manufacturer specifications. Plates were centrifuged after a 2-hour incubation, then the TR-FRET assay was read in the EnVision (Perkin Elmer, Waltham, MA). The original plate containing the cells and pre-treated with compounds for 72 hours was used to do the Luminescence part of the assay. 5 uL of Promega (Madison, WI) CellTiter-Glo was added to this plate and centrifuged. Plates were incubated for 10 minutes, then a luminescence read was performed on a ViewLux reader (Perkin Elmer, Waltham, MA).
Results and discussion
Maintenance log
Initial deployment and usage of the orbital shakers revealed a few components that were prone to failure, especially when run 24 hours a day 7 days a week in a high humidity environment. Once these failures became evident, and in an effort to address troubleshooting and repair efforts, a simple maintenance log was created to document whether repairs were effective. Each orbital shaker was assigned a unique identifier so that issues with individual units could be tracked over time if needed. When failures were observed, the failed parts would be modified if needed, with the relevant CAD model updated to prevent future repeat failures and a new part 3D printed and installed. Based on this experience, we expect that future projects of a similar nature will be best served by establishing maintenance logs to track failures from the onset of the project.
The modes of failure seen most frequently were:
•
3D printed motor coupler part failure. We found that the 3D printed part often failed along 3D printed layer interfaces. We resolved this by increasing the part thickness and 3D printing in an orientation that resulted in the tension load being 45 degrees to layers instead of normal to layers [
Motor coupler and shaker platform screws coming loose. We resolved this by applying Loctite Blue 242 Threadlocker to all screws which interfaced with nuts or heat-set inserts.
•
Bearings coming loose from 3D printed components. The initial design of the DIYbio orbital shaker had all five bearings installed in their respective 3D printed components with hot glue. Over time, this glue would degrade and break loose, resulting in loose bearings and damaged components. We resolved this by redesigning the 3D printed components allow the bearings to be press fit into the components rather than held in place with adhesives.
•
Nuts coming loose from 3D printed components. Assembly of components with nuts required either use of adhesive or holding the fastener components in place while simultaneously completing the assembly operations. Further, many of these nuts were installed on the bottom of the 3D printed parts, meaning they had to be manually held in place in difficult to reach locations or orientations while fasteners were installed. For those nuts held in place with adhesive, over time the adhesive would fail, resulting in parts coming loose, or nuts would fall away from the parts during subsequent disassembly for repair work due to the nuts being installed on the bottom of the parts. We resolved this by replacing all nuts with threaded heat-set inserts that could be installed into 3D printed parts with a soldering iron [
Power plug coming loose from enclosure. The glue holding the power adapter plug in place would degrade, causing the plug to recess into the shaker and not connect with the power supply connector. We modified the design for the case was to have a press fit installation for the power adapter plug to address this issue.
After the items mentioned above were addressed and the shakers were proven to be robust via testing, the shakers were utilized for assay development and a full-scale HTS campaign. To prevent shakers from moving around inside the incubator while in use, all shakers were installed in place on the incubator shelving with double-sided adhesive tape. If additional issues occur in the future and become more time-intensive to resolve, it may be prudent to utilize maintenance log data to perform a more formal analysis of the failures, such as a Pareto analysis.
Screening assay results
A large-scale screen was run, comprised of 640,011 compounds from the Scripps Drug Discovery Library. The cells were grown and expanded using the orbital shakers. 60 plates per day were successfully plated to perform the screening for the two modes across multiple days. This multiplex assay consisted of two endpoint readout technologies: luminescence and time-resolved fluorescence energy transfer (TR-FRET). Both assay readouts were accessed on the same day with assay statistics. Assay plates were determined acceptable only if Z' > 0.5 [
]. Z' for the cell viability luminescence assay was 0.88 +/- 0.02 with an average signal to base level signal of 19.24 +/- 5.75 and for the TR-FRET mode, Z' was 0.73 +/- 0.04 with an average signal to base level signal of 8.45 +/- 1.50.
Future improvements
After using the redesigned orbital shakers for 2 years, it was observed that when multiple units are stacked closely together on an incubator shelf, it becomes difficult to reach behind the shaker to plug or unplug the power cable. Additionally, it is also difficult to reach the power switch on the left side of the shaker. As such, the CAD model has been updated to a third version (v3) of the design so that future shakers will have the speed control, power switch, and power plug all accessible from the front of the unit. Existing units will be updated to this new design when maintenance or repair is needed. Initial deployment of this updated v3 design can be seen in Fig. 5.
Fig. 5Orbital shaker repair kit with labelled compartments (top) and spare parts (bottom) visible.
Due to the ongoing maintenance requirements described in the previous section, we have also assembled an orbital shaker repair kit, seen in Fig. 6, which contains all parts needed to repair an orbital shaker or construct a new orbital shaker, except for those parts which are too large to fit in the kit.
Fig. 6Second version of the DIYbio orbital shaker (left) versus third version (right).
Future iterations of the orbital shaker may also be updated to include an LCD display to report the current shaker speed in revolutions per minute (rpm). Future iterations could also include improvements to make assembly of the control electronics easier. This could include utilizing commercially available stackable shield-style components similar to those seen in the Arduino ecosystem or adding screw-terminal connectors to limit the amount of soldering needed for assembly. For these changes, the footprint of the shaker enclosure may need to be increased if such components had a larger footprint than the custom PCB currently used.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Competing Interest
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
00003-6&id=gr1.jpg){kind=link}
00003-6&id=gr2.jpg){kind=link}
00003-6&id=gr3.jpg){kind=link}
00003-6&id=gr4.jpg){kind=link}
00003-6&id=gr5.jpg){kind=link}
00003-6&id=gr6.jpg){kind=link}