Enterprise AR Solutions: Build or Buy?

Build or buy? This is a question with no simple answers. As enterprises contemplate deploying AR solutions, one of the first questions to confront them is a fundamental one: should we build or buy? AREA Researcher Dr. Michael Rygol explores the factors that may help organizations answer this critical question. Please visit thearea.org.

Episode Transcript

(00:01):
Hello,
I'm Michael Rygol and welcome to thelatest in this series of AREA podcasts.
Today's podcast discusses the questionfaced by many enterprises when
considering an augmented realitystrategy, should we build or buy?
This is a question with nosimple answers as we'll see.
The build or buy decision essentiallyboils down to determining the relative

(00:23):
priorities of cost control,solution availability,
and time to market intraditional solution deployments.
The advantages and disadvantages of eachapproach can be summarized as follows.
Some reasons to build a customsolution include increased control,
which provides the ability for you todevelop a solution that exactly meets your

(00:43):
process and market differentiationneeds a potentially lower through life
deployment cost and the ability toexercise full control over future
development roadmaps. However,
there are some reasons tonot build a custom solution,
lack of focus or domainexpertise within your business,
or even a shortage of available skills.It could require significant

(01:06):
investment in tools, processes, andstaff to specify, architect, develop,
document, manage, andsupport the solution,
whether externally orinternally developed.
Building a custom solutionpotentially has higher levels of risk,
of cost overrun and failure to meetproduct objectives. Conversely,
if we consider the reasons to purchasea commercially available solution,

(01:28):
these include there's adecreased time to value.
This solution is potentially availableright now, economies of scale.
The vendor of the solution is clearlysolving this problem for numerous other
companies and can leveragetheir domain expertise. Also,
the vendor likely has a proven abilityto solve potentially difficult technical

(01:49):
challenges,
but bear in mind that there are alsoreasons to consider avoiding purchasing a
commercial solution.
You may be vulnerable to vendor lock-inor longer term viability of the vendor.
Their solution may not bean exact fit to your needs,
which may be difficult or costly toovercome with customizations. Depending

(02:10):
on the nature of the contractyou have with the vendor,
you may be subject to ongoing andpotentially unpredictable payments to that
vendor. When it comes toenterprise AR deployment,
the build or buy deliberations needto take in additional considerations.
These may include some or all of thefollowing implementations on novel or
unfamiliar hardware development ofadvanced computer vision capabilities,

(02:35):
application development based onAR toolkits or data processing,
protection and optimization. For example,
for 3D models or integrationinto enterprise systems
development of custom content for newmethods of deployment and user interaction
or customization of the basesolution to meet specific needs.

(02:56):
There may also be wider implications ifsignificant Content must be created by
internal or external personnel.
A typical set of steps leading to thedecision-making phase include identify the
business use case, perform investmentanalysis, and secure the budget,
define weighted requirements for thesolution for the identified business

(03:17):
problems or opportunities,
identify potential vendors andtheir commercial solutions.
Then perform a gap analysis betweencommercial offerings and solution
requirements.
Then identify whether gaps can beclosed with customization or custom
development. After that,
then we can perform a cost analysis ofthe internal or external development

(03:40):
versus the offerings madeby the commercial solution and finally evaluate these
options and make a strategy decision.
It's important to understandthat enterprise AR based solution needs may vary
significantly according to thetarget use case. For example,
the AR enhanced remote assistance usecase likely requires low levels of

(04:00):
customization or new contentto be developed. However,
use cases such as sales demonstratorsor design review tools or AR
driven service technician procedureswill likely require increasing levels of
system integration,
custom content to be developedor reused and process automation,
all of which may result inthe out-of-the box offering,

(04:22):
presenting more significant gaps tobe closed prior to deployment. Such
factors may play an important influencein the build or by decision-making
process. The AI enhanced product salesdemonstrator use case, for example,
will likely require lowerlevels of integration with business systems and data,
and therefore coupled with other factorssuch as time criticality and reduced

(04:45):
longevity needs may make it appropriateto subcontract all software development
and content creation to a thirdparty. If your use case is unusual,
then you may need to consider purchasingan AR platform that allows custom
development, whether via drag and dropordering, coding, or other mechanisms.
Typical questions to consider when youare making the build or buy decision are

(05:08):
as follows.
Have you identified the businessapplications or problems to be solved?
Have you developed the requirementsneeded to address the business problem?
Are there commercial offerings claimingto provide a solution for your use
cases?
Are you confident that the solutionmeets your functional requirements? Would
more than one commercial productbe needed to provide the solution?

(05:28):
Are you confident of the solutionprovider's financial viability?
Other more AR specificconsiderations include,
does the offering meet yoursecurity and privacy policies?
Does the solution offeracceptable performance?
Does the software support yourtarget hardware? Additionally,
does the solution offer integrationcapabilities with your enterprise systems

(05:51):
such as IOT or PLM orERP? And importantly,
how closely does the offering meetyour defined requirements? Also,
what abilities are there to allow reuseof existing data assets rather than
creating new content fromscratch? If there are gaps,
can the solution be customized tomeet your exact needs? And finally,

(06:13):
does the solution meet yourenterprise scalability objectives?
To conclude this editorial has exploreda number of considerations that are
important when seeking to adoptAR. In an enterprise setting,
companies may be tempted to developprototype applications when first
investigating ar,
perhaps using one or more of thecommercially available toolkits. However,

(06:36):
there are clearly a number of importantaspects to consider in reaching a build
or buy decision.
It's unlikely that an industrial companywill develop an in-house AR application
from the ground up as this requiressignificant expertise in numerous areas
including computer vision, 3D, computergraphics, mobile device management,
and more.

(06:57):
If your use case is truly unique andthere are no commercial products that
support your use case,
then your only option may be to developthe solution this way. Far more likely,
however, is the decision to purchasea commercial off the shelf solution.
As we've discussed, and dependingupon your target use case,
there may be significant requirementson systems integration, data processing,

(07:21):
content creation,
and other forms of customization requiredprior to considering a deployable
solution. As mentioned earlier,
the decision is often driven by therequirements of cost control and timing.
If cost and timing are a higher priority,
then a commercial offering is likelythe more appropriate solution.

(07:41):
If control is most important,
then it's perhaps better topursue internal development or more likely contract
work to a third party. Ultimately,the decision is yours. However,
prior to making that decision,
we recommend that you look at theofferings of the area solution provider
members who I'm sure will be happyto discuss and hopefully meet your

(08:02):
requirements. Thanks for listening.