Content uploaded by Joshua M Pearce
Author content
All content in this area was uploaded by Joshua M Pearce on Nov 08, 2015
Content may be subject to copyright.
To be published: Joshua M. Pearce. Return on Investment for Open Source Hardware Development. Science and Public Policy. DOI :10.1093/scipol/scv034
Return on Investment for Open Source Hardware Development
J.M. Pearce a*
a. Department of Materials Science & Engineering and Department of Electrical & Computer
Engineering, Michigan Technological University, MI, 49931, USA
* corresponding author: 601 M&M Building, 1400 Townsend Drive, Houghton, MI 49931-1295
pearce@mtu.edu
Abstract
The availability of free and open source hardware designs that can be replicated with low-cost
3-D printers provide large values to scientists that need highly-customized low-volume production
scientific equipment. Digital manufacturing technologies have only recently become widespread and
the return on investment (ROI) was not clear, so funding for open hardware development was
historically sparse. This paper clarifies a method for determining an ROI for FOSH scientific hardware
development. By using open source hardware design that can be manufactured digitally the relatively
minor development costs result in enormous ROIs for the scientific community. A case study is
presented of an syringe pump released under open-licenses, which results in ROIs for funders ranging
from 100s to 1,000s of percent after only a few months. It is clear that policies encouraging free and
open source scientific hardware development should be made by organizations interested in maximizing
return on public investments for science.
Introduction
As prominent voices in the scientific research community
have pointed out, funding cuts continue to reduce grant
success rates [1,2]. This creates hyper-competitiveness with
the concomitant diminishing of risk taking and innovation
among researchers of all ages and desertion of many young
investigators [1,2]. In addition, researchers consume an
ever-increasing fraction of their time with grant writing
rather than actually doing science. Simultaneously, the
national average overhead charged on grants (indirect costs
primarily used to subsidize administrative salaries and
building depreciation), has climbed to 52% [3], which
further limits scientists' ability to do research with the
hard-earned funding they do obtain. Experimentalists are
perhaps the hardest hit as low-volume, highly-specialized
equipment needed to push further scientific progress
continues to demand premium and often shockingly high
prices on the market.
Fortunately, advances in low-cost electronics and 3-D
printing [4] enables a new paradigm of scientific
equipment production where scientists in the developed
and developing worlds can fabricate tools themselves from
digital plans [5-9]. There has been an exponential rise in
designs for hardware released under open-source, creative
commons licenses or placed in the public domain [10]
including the 3-D printers themselves [4]. This free and
open source hardware (FOSH or open hardware) is shared
between scientists by providing the bill of materials,
schematics, assembly instructions, and procedures needed
to fabricate a digital replica of the original [9]. FOSH
reduces redundant problem solving in laboratories around
the world, accelerates innovation due to rapid laterally-
scaled feedback [5-9]. In this way hardware can benefit
from the same methods that have proven so successful with
free and open source software [11-12]. The number and
variety of FOSH scientific tools is rapidly expanding
[6,9,13-20]. For scientists that need access to highly-
customized low-volume tools the open source method and
digital replication can result in significant cost savings
[6,9,13]. Although research capital goes farther with
digitally distributed open designs, it still costs money to
develop them. This paper investigates the return on
investment (ROI) possible for fund agencies such as the
NIH or NSF for developing open hardware for science. A
brief case study is provided to illustrate the economic
calculations and then policy recommendations are made to
maximize the ROI in research and development in any field
accessible to open hardware design.
1
To be published: Joshua M. Pearce. Return on Investment for Open Source Hardware Development. Science and Public Policy. DOI :10.1093/scipol/scv034
Calculations
The value obtained from a FOSH design can be determined
from the downloaded substitution valuation [21] (VD) at a
given time based on the number of downloads (ND) for a
given design at time, t:
VD
(
t
)
=
(
Cp−Cf
)
×P×ND
(
t
)
(1)
Where Cp is the cost to purchase a traditionally
manufactured product, Cf is the marginal cost to fabricate it
digitally (e.g. open-source 3-D printing), and P is the
percent of downloads resulting in a product. It should be
pointed out that P is subject to error as downloading a
design does not guarantee manufacturing. On the other
hand, a single download could be fabricated many times,
traded via email, memory stick or posted on P2P websites
that are beyond conventional tracking. The savings are
maximized for custom low-volume scientific equipment
where Cf is generally only 1-10% of Cp [9,17], creating a
90-99% savings.
Investment in scientific open source hardware can thus
create a return on investment that can be calculated by:
ROI= VD
(
t
)
−I
I
(2)
where I is the cost of the investment in the development of
the FOSH scientific tool.
Case Study
Consider the case study of a simple open-source syringe
pump library design [22], which may be government
funded for scientific innovation acceleration, but also have
applications in STEM education and medicine. The low-
cost open source pumps are completely customizable
allowing both the volume and the motor to scale for
specific applications such as any research activity
including carefully controlled dosing of reagents,
pharmaceuticals, and other applications. The design, bill of
materials and assembly instructions are globally available
to anyone wishing to use them.
The pump library was designed using open-source and
freely available OpenSCAD, which is script-based,
parametric CAD package enabling scientists to customize
the design for themselves [22]. The majority of the pump
parts can be fabricated with an open-source RepRap 3-D
printer and readily available parts such as a stepper motor
and steel rods. The pumps can be used as wireless control
devices attached to an open-source Raspberry Pi computer.
Performance of the syringe pumps generated by the open
source library were found to be consistent with the quality
of commercial syringe pumps [22].
As of this writing (Feb. 2015) the designs for the open
source pump, which were released in Sept. 2014, have been
downloaded from two digital repositories a total of ND =
1035 times (224 on Thingiverse [23] and 811 on
Youmagine [24]). The cost to purchase a traditionally
manufactured syringe pump, Cp, ranges from $260-$1,509
for a single pump and $1,800-$2606 for a dual pump [22].
Cf for the materials for an open source syringe pump is $97
and $154 for the single and double pump, respectively [22].
The time to assemble either the single or double pump is
less than an hour and can be accomplished by a non-expert.
Although the time to print the components is less than four
hours on a conventional RepRap, workers can do other
tasks while printing. The assembler hourly rate is assumed
to be $10/hour because no special skills are needed. P was
assumed to be 1 as informal discussions with 50 RepRap
owners found that the vast majority of designs downloaded
were printed. This is because although RepRap owners
may view many designs they only download the STL files
(STereoLithography is a file format used for 3-D printers)
of the designs they intend to print. This provides a savings
for substituting the open source pump for a commercial one
of $153 to $1,402 for a single and $1,636 to $2,442 for the
double pump. Thus, following equation 1 the value VDT of
the pump library on Feb. 5, 2014 ranged from over
$168,000 to over $2.5 million.
The investment needed to create the pump library was
trivial in comparison to the savings. The mechanical
designs were completed by experienced engineers in <6
worker-hours. To print and revise the five 3-D printed
components took 3 hours, assembly less than 1 hour and
software development and Pi wiring less than 16 hours.
The total design and prototyping time was less than 26
hours total. This schedule was possible because the
designers were experienced with similar designs and had
access to all of the components. Let us assume that the
pump design was completed as part of an NSF or NIH
grant that involved overhead, materials, validation testing
from a PhD student over a semester thus involving an
initial investment, I, of $30,000. By equation 2 this
provides ROIs ranging from 460% to over 8,300%.
These ROIs, however, are conservative as the value of the
FOSH in labs is not only the amount saved for research
directly, but also includes the value of the overhead
(indirect costs) charged on grants to purchase the product
2
To be published: Joshua M. Pearce. Return on Investment for Open Source Hardware Development. Science and Public Policy. DOI :10.1093/scipol/scv034
commercially. When the national average overhead rate of
52% is included on the cost of conventional equipment the
ROIs increase to range from over 750% to over 12,000%,
respectively.
Discussion
Some care must be taken in evaluating these ROIs. First, to
be conservative it should be pointed out that although the
majority of down loaders can be presumed to be American,
not all of the savings would accrue to those that U.S.
government funding agencies are supporting. Although,
improving science globally will help researchers in the U.S.
indirectly it may not be viewed as a direct ROI. On the
other hand, the low end value is from a simple infusion
pump with considerably less functionality than the FOSH
syringe pumps. It is a safe assumption that the majority of
down loaders are likely to be replacing more sophisticated
scientific devices. In addition, it also is clear that the
number of downloads and fabrications of the syringe
pumps will continue to expand with time providing ever
increasing value for the scientific community and higher
ROIs for funding such development.
There are other benefits as well. As the pump meets the
standards for research and has already been vetted it seems
reasonable that it would be most likely to be adopted
university labs first. This would results in potential
additional value for both scientific research and education.
By decreasing the costs of research equipment more of the
diminishing resources for science are available to do
science. For example, if a quad syringe pump is fabricated
for a molecular biology lab the savings would be enough to
hire a summer undergraduate researcher – thus presumably
increasing the scientific discovery rate. In addition, as the
designs are FOSH, other labs are already improving the
tool and re-sharing the improvements (e.g. [25]). This
creates superior scientific equipment in the future with no
additional expenditures. Similarly, due to lower costs,
FOSH could be used in classroom or lab courses at every
level, thus improving education.
Quantifying the value of an increased rate of discovery in
science and medicine because of lower costs or superior
equipment or better education entails specific detailed
studies. These studies, for example, could utilize follow-up
surveys targeted at all down loaders of the design and STL
files to get a precise account of P, Cp and Cf. Then, in turn
these initial calculated ROIs could be augmented by
following the results of groups/schools that adopted the
open hardware. This would be useful for obtaining exact
values of ROI. However, the methodology used here
provides reasonably accurate values based on conservative
assumptions of simply replacing proprietary equipment.
Qualitatively, it is clear that free and open source scientific
hardware has ROIs set by the minimum calculated by
equations 1 and 2 and these are significantly higher than
those involving proprietary investment by public funding
groups.
It should also be noted that as free and open source
hardware becomes more commonplace in the scientific
establishment there may be a cultural shift within labs to
employ 'makers' to build and troubleshoot open source
equipment. The 'maker movement' is a growing culture of
hands-on making, creating, designing, and innovating at all
ages [26,27]. Such troubleshooting is already provided in a
limited way by technicians, research assistant students and
research scientists, but would be expected to expand and
provide positions for the growing number of makers as
more scientific hardware becomes completely accessible
and able to be customized.
Finally, it should be pointed out that there is nothing at all
remarkable about the syringe pump used as an example
here. Any one of dozens of other existing open source
scientific hardware designs that have a Cf that is a tiny
fraction of Cp [9,17] and that have been downloaded a small
number of times would show similar enormous ROIs. In
addition, there are numerous types of equipment that could
be redesigned as FOSH with low fabrication costs that
could be widely applied outside of scientific labs (e.g. hand
held nitrate testers for water quality or use for farmers to
determine fertilizer requirements) that would have
enormous potential returns for society.
For entities such as the NIH, which are trying to leverage
their investments for the greatest common good, FOSH
development has extremely high potential ROIs.
Traditional investments in proprietary development would
not come close to such a return, nor would the source be
available for others to immediately begin working on
because of the 20-year external innovation hiatus
demanded by the current patent system [28].
It is clear that federal funding (such as through the NIH,
NSF, DOE, DOA, DOD, and NASA, etc.) should be
prioritized for the development of open-source scientific
hardware because of the enormous potential ROIs for the
nation's scientific community. This can be accomplished
with a combination of traditional CFPs for academic grants
and programs like SBIR and STTR programs. In addition,
3
To be published: Joshua M. Pearce. Return on Investment for Open Source Hardware Development. Science and Public Policy. DOI :10.1093/scipol/scv034
the U.S. can run national contests like the X-prize or “first
to make” specific technical goal “bounties”.
Just as proprietary tools on the market are, all FOSH
scientific designs should be vetted, tested and validated.
This will largely eliminate the technical risks for labs to
adopt the use of the hardware, while at the same time
ensuring that scientific equipment no longer becomes
obsolete as proprietary systems can when a company loses
key personnel, discontinues a product line or goes out of
business. Funding for this validation would be the majority
of the cost of scientific FOSH. Then, in order to ensure the
widest distribution of the free technologies, NIH's 3D Print
Exchange should be expanded to act as a national free on-
line catalog of tested, vetted and validated free and open-
source scientific hardware. The NIH's 3D Print Exchange
should be augmented to be able to house the bill of
materials, digital designs, instructions for assembly and
operation, and the source code for all software and
firmware.
Conclusions
By funding open source hardware design that can be
manufactured digitally, the relatively minor development
costs results in enormous ROIs for the scientific
community. As the designs are reusable, with solid
modeling and 3-D printing, designs can be expanded or
joined together rapidly increasing the rate of innovation.
The case study presented here found ROIs of 100s to
1000s of percent from a relatively simple scientific device
being released under open-licenses. It is clear that free and
open source scientific hardware development should be
funded by organizations interested in maximizing return on
public investments.
Funding
This work was supported by the National Science
Foundation [NSF STTR Phase I Grant #IIP-1417061].
Conflict of Interest
None
References
[1] ‘No money, no research’. (2014) Nature Methods 11:
1077–1077.
[2] Alberts, B., Kirschner, M.W., Tilghman, S., Varmus, H.
(2014) 'Rescuing US biomedical research from its systemic
flaws'. Proceedings of the National Academy of Science
USA 111: 5773–5777.
[3] Jan T. (2013) 'Research giants win on federal funding'.
Boston Globe. March 18, 2013.
[4] Jones, R., Haufe, P., Sells, E. (2011) ‘RepRap – the
replicating prototyper’. Robotica. 29/1:177–191.
[5] Jones, N. (2012) ‘Science in three dimensions: the print
revolution’. Nature 487: 22–23.
[6] Pearce, J.M. (2012) ‘Building Research Equipment with
Free, Open-Source Hardware’. Science 337/6100: 1303-
1304.
[7] ‘Sharing blueprints for better research’. (2013) Nature
Methods 10: 591–591.
[8] Pearce, J.M., Blair, C., Laciak, K., Andrews, R., Nosrat,
A., Zelenika-Zovko, I. (2010) ‘3-D printing of open source
appropriate technologies for self-directed sustainable
development’. Journal of Sustainable Development 3/4:17-
29.
[9] Pearce, J.M. (2014) ‘Open-Source Lab: How to Build
Your Own Hardware and Reduce Research Costs’.
Elsevier: New York.
[10] Wittbrodt, B., Glover, A., Laureto, J., Anzalone, G.,
Opplinger, D., Irwin, J., Pearce, J. (2013) ‘Life-cycle
economic analysis of distributed manufacturing with open-
source 3-D printers’. Mechatronics 23/6:713-726.
[11] Ball, P. (2003) ‘Openness makes software better
sooner’. Nature, 030623-6.
[12] Lakhani, K.R., Hippel, E.V. (2003) ‘How open source
software works: free user-to-user assistance’. Research
Policy 32:923- 943.
[13] Pitrone, P. G., Schindelin, J., Stuyvenberg, L.,
Preibisch, S., Weber, M., Eliceiri, K., Huisken, J.,
Tomancak, P. (2013) ‘OpenSPIM: an open-access light-
sheet microscopy platform’. Nature Methods 10: 598–599.
[14] Gualda, E. J., Vale, T., Almada, P., Feijó, J., Martins,
G., Moreno, N. (2013) ‘OpenSpinMicroscopy: an open-
source integrated microscopy platform’. Nature Methods
10: 599–600.
[15] Carvalho, M. C., Eyre, B. D. (2013) ‘A low cost, easy
to build, portable and universal autosampler for liquids’.
Methods in Oceanography 8:23-32.
[16] Wolf, L.K. (2013) ‘Print-A-Lab’. Chem & Eng News
91: 44-45.
[17] Zhang, C., Pearce, J., Faria, R., Anzalone, N. (2013)
‘Open-Source 3D-Printable Optics Equipment’. PLoS ONE
8/3, e59840.
[18] Herrmann, K. H., Gartner, C., Gullmar, D., Kramer,
M., Reichenbach, J.R. (2014) ‘3D printing of MRI
compatible components: why every MRI research group
should have a low-budget 3D printer’. Med Eng & Physics
36/10: 1373–1380.
[19] Gross, B. C., Erkal, J., Lockwood, S., Chen, C.,
Spence, D. (2014) ‘Evaluation of 3D Printing and Its
4
To be published: Joshua M. Pearce. Return on Investment for Open Source Hardware Development. Science and Public Policy. DOI :10.1093/scipol/scv034
Potential Impact on Biotechnology and the Chemical
Sciences’. Analytical Chem 86/7: 3240-3253.
[20] Chiu, S. H., Urban, P. L. (2015) ‘Compact 3D-printed
interface for coupling open digital microchips with Venturi
easy ambient sonic-spray ionization mass spectrometry’.
Biosensors and Bioelectronics. 64: 260–268.
[21] Pearce, J.M. (2015) ‘Quantifying the Value of Open
Source Hardware Development’. Modern Economy 6:1-11.
[22] Wijnen, B., Hunt, E., Anzalone, G., Pearce, J. (2014)
‘Open-Source Syringe Pump Library’. PLoS ONE 9/9,
e107216.
[23] Pearce, J. ‘Open Source Syringe Pump’. (2014)
Available at: http://www.thingiverse.com/thing:475841
(Accessed: 5th February 2014)
[24] Pearce, J. ‘Syringe Pump’ (2014) Available at:
https://www.youmagine.com/designs/syringe-pump
(Accessed: 5th February 2014).
[25] Lynch, D. Modified Open Source Syringe Pump Parts
(2014). Available at:
http://www.thingiverse.com/thing:578010 (Accessed: 5th
February 2014)
[26] Dougherty, D. (2012). 'The maker movement.'
Innovations, 7(3), 11-14.
[27] Peppler, K., Bender, S. (2013) 'Maker movement
spreads innovation one project at a time'. Phi Delta
Kappan, 95(3): 22-27.
[28] Pearce, J. M. (2012) ‘Make nanotechnology research
open-source’. Nature 491: 519–521.
5
