Cloud Computing in Technical Communication

Hardly any other topic has captivated the IT sector to this great an extent since the CeBIT 2011, as has cloud computing. According to the Experton Group, revenues of over Euro 1.1 billion were invested in technologies, services and consulting related to cloud computing in Germany in 2010 [1]. The research company expects an average annual growth of 40 percent till 2015.

A definition of cloud computing

The National Institute of Standards and Technology, NIST, defines cloud computing as “a model for smooth network access available on call to a common pool of configurable computing resources (e. g. networks, servers, memory, applications and services) that can be provided quickly and released with low management effort and minimal intervention by the service provider” [2]. The NIST definition details the attributes that the cloud solution must possess, and the basic models of use and significant services that can be realised:

• Large bandwidth for access channels – different end devices such as PCs, mobile telephones, PDAs or laptops have access to the required resources over the network.
• Rapid scalability – the resources are provided flexibly similar to electrical power. This means, they can be scaled efficiently, in the ideal case automatically, corresponding to the requirements of the applications.
• Measurable services – it must be possible to record the services quantitatively, so that providers and users obtain complete transparency and a usage based billing.
• Order as needed – the resources can be ordered by the user independently and as required.
• Resource pooling – the resources are divided in several data pools and several clients are provided.

Three basic categories have emerged from the services:

• “Infrastructure as a Service” (IaaS) provides infrastructural resources, such as computing power and memory.
• “Platform as a Service” (PaaS)-services are primarily addressed to developers and provide development as well as operating environments for applications.
• The model “Software as a Service” (SaaS) is probably best known. Specific complete applications are made available here.

Cloud solutions can be classified under four models of use:

• A private cloud is operated solely for one organization, e.g. for a company.
• A public cloud is accessible to the public.
• Hybrid cloud models are combinations of different cloud forms.
• We speak of community clouds when different organizations with similar interests share the use of infrastructural resources, e.g. associations.

Constraints for implementation

In a current study of the telecommunication service provider COLT, seven of ten technical leaders mention security as their greatest concern when implementing cloud services [3]. Without question it needs some getting used to, to know that internal data is not lying on the company server, but in an external data center. Critics ask how data security – keyword “Backup und Recovery” – and the protection from manipulation or unauthorised access is managed. However, these doubts come to bear in case of other outsourcing projects as well.

To reduces risks, the user should clarify:

• where the data lies physically
• which legal area the data exists in
• what authentication mechanisms are applied
• whether the data is transferred over secured connections  – https protocol
• how high the stability of the hardware is
• what the back up concept looks like
• what data is to be managed in the cloud solution
• whether special data protection regulations are to be considered, e.g. in relation to personal data
• whether special contractual provisions are requried.

Tips for selecting a cloud solution

Data security and data protection: Usually, data centers are equipped with significantly  greater security measures as compared to other companies. Providers should give comprehensive information about the measures they have taken for data protection and security. Customers often request a copy of their data, which is however linked to additional costs and contradicts the basic concept of cloud computing. It is better when the provider can exactly prove that he backs up/secures the data.

Dependence on service provider: There is then the question of dependence on the cloud provider. Therefore, it should be clarified whether the cloud solution can be reverted to the internal IT environment and what the migration costs would be. The definition of service level agreements, SLA, and the scalability of the application and the infrastructure are additional factors for the performance/efficiency and availability of the cloud, even when requirements grow.

Performance/efficiency and availability: It is important for the response time of the software to be short for high performance work within the technical documentation department. Ideally, the user should not be able to identify any difference whether the server exists within the company or in a data center of the IT provider. It should be possible to test the response time of the software in a real environment in advance.

Support and service quality: User satisfaction rises and falls not just with the scope of functions of the software but also with the service quality of the service provider. Long queues or unsatisfactory responses reduce acceptance. It is best when a direct and competent contact person is asigned to the user and takes care of the questions and problems.

Cloud for technical documentation

Cloud solutions are possible or already available for most applications that have to do with technical documentation. This includes editing systems, as well as authoring systems, translation memory systems and terminology databases.

If a company implements the cloud model for its technical documentation, the architecture could look like this: The editing system, author supporting software and the terminology database are installed on the central servers. These servers are maintained by the service provider. He manages the applications and takes care of the update of services. The users at the end customer, e.g. the technical editors, use the services from their end device via internet browser. Even partners, such as external editing service providers or translators have access, which may make sense, especially for common population and use of the terminology database of a translation memory system for instance. All functions can thus be steered over the internet; even the data management is centralized.

Unlike systems based on client and server, additional software on the client computer is not required for working in the cloud. Different locations, external service providers, partners and customer can thus be integrated.

This model can be helpful for supply documentation as well. For instance, if an automotive manufacturer creates the documentation for his vehicles in a web based system, then the documents of the supplier can also be integrated there. The supplier gets a corresponding input right. Other software to be installed by him is not required

What are the advantages?

What are the differences between implementing technical documentation with the help of cloud computing as compared to classical client/server based implementation? For example, implementing a new technology can be cost effective. For small and mid-size companies that desist from acquiring new systems for cost reasons, cloud computing can be an impetus. Translation processes and terminology checks integrated in cloud computing reduce costs and enable a more flexible application of technical documentation.

Furthermore, the user is not tied down to a specific system provider. He can usually terminate the services within a short period. Here it is important that the data is available in XML based form. Then it can be migrated quicker in case of a return to own company or in case of change of provider.

The risks in case of data security can be minimized by a security strategies and backup strategies. It is important to maintain updated security standards through periodic reviews.

Setting up new users

One advantage of cloud computing lies in the flexibility and scalability. Additional users are enabled to work in a short time, without requiring any installations on the user side. There is no long term rigid binding to user licences. These can be added or terminated again as required. The same applies to optional services and functionalities. A byproduct of implementing cloud solutions is that the company can concentrate on its core business and not on IT projects.

Presently there is a lot to be said in favor of private and hybrid cloud models with respect to usage models. The data security is in one’s own hand, the performance/efficiency can be governed by the user himself and the use as per own specifications is simpler to implement. However, public and community clouds can also become interesting with increasing standardization. XML offers a basis as an exchange format.

Layouts can be used by several companies, at least to some extent.  The scalability primarily saves costs in multi client enabled cloud solutions, since these are distributed over several participants. It is necessary to reflect on the extent to which a common standard like the DITA, PI-Mod or docbook can be used for the information model.

Summary

Moving technical documentation to a cloud has advantages. The user can work productively at a comparatively low expense and with comparatively less effort and need not bother about software and updates.  If other locations want to work with the system, they are simply enabled. Therefore cloud computing is a solution especially for companies with distributed locations.

References

[1]    Experton Group. (2010)

[2]    Mell, P./Grance, T. (2011): NIST Special Publication and Reference Materials.

[3]    Colt Technology Services GmbH (2010)