Why should CSU Provision Devices?
The mobile device market is extremely suitable to a BYOD model but it comes with significant risks and implications for institutional resources. In particular, there will be a requirement for increased support simply because of the dispersed and diverse profile of devices available. This diversity is only set to increase in the next 12 months as new operating systems, FirefoxOS, Ubuntu Touch and Tizen, are being rolled out. This is on top of the significant fragmentation of Android operating systems and massive range of devices with wide ranging capabilities.
This diversity creates significant issues for the development of specific solutions for mobile devices and impacts the ability of the university to provide best of breed solutions instead of low-common denominator answers. In terms of developing apps and delivering content, it presents significant challenges as there is no ability to develop a one size fits solution, instead requiring a range of solutions, which increases cost and time for development.
Provisioning devices means taking a single vendor approach, not necessarily a single device approach, but one where there is a common operating system, hardware specifications and device types. A single vendor approach comes with the risk of lock-in, but also has significant benefits for the university:
- Reduction in support requirements as there are “Known knowns and known unknowns” improving risk management, support resources and issues can be replicated and responded to in a timely manner.
- An organic community of practice can form around a common ecosystem capable of supporting each other with regard to practices and techniques reducing the need for direct intervention.
- Simplification of development and testing due to platform consistency applies to the development of content, resources, applications and systems.
- Devices can be managed to a greater degree and open new opportunities for different types of ownership models, such as leased, loaned or contracted
From initial discussions in the Project the implementation of a range of provisioning models addressed concerns around costs and sustainability, and provide the institution with a best practice model. It could also provide a way for CSU to leverage its relationships with corporate interests to develop some new, innovative methods ensuring our students and staff are provided with cutting edge, high quality equipment.
Some example models could include:
Institution Provided – CSU would provide the device to staff and students to own and cover costs, or recover through other measures. This would only be suitable in small numbers but could be used to provide assistance to Indigenous and low SES students.
Institutional Loans – CSU would provide the device as a loan to staff or students. These could be similar to leasing arrangements and made contingent on employment and enrolment. The institution would need to make the large initial investment but would be able to recoup costs. This would lessen the upfront impact on students, faculties and divisions.
Institution Sourced – CSU would act as an agent for various hardware providers and leverage bulk buying and discounting to reduce costs. Existing internal services could be expanded to students who would benefit from savings.
Vendor Sourced – CSU would establish working relationships with vendors to provide staff and students with discounts, with all sales conducted outside the university. This would require negotiations with external manufacturers and/or vendors, but could result in significant savings without major internal investment.
Data Contracts – CSU could negotiate an agreement with a telecommunications company to provision devices to staff and students. Each contract could operate outside the university but leverage the bulk buying capabilities to reduce costs to students. This arrangement could extend to allow CSU data to be un-metered, in effect providing students and staff with free data for their work, learning and teaching. This method could leverage on the type of margins that retail sellers earn from sales, but could be used to offset costs to students.
Along with choosing the provisioning model, there may also be scope to investigate deployment models. Most tablet solutions can support managed and unmanaged deployment models, providing flexible solutions to individuals and institutions. Each method has specific pros and cons that need to be matched to the specific purpose and goal of any deployment to ensure suitability.
The processes around Managed Devices are similar to current methods used in DIT. The university would retain ownership and management control of the device and would also need purchase software to be made available. This method requires a manual initialisation process where an ‘image’ is created, and then rolled out across the other devices. The image can contain protocols to simplify the setup of some processes like email, restrict functions and features on the device, allow tracking to ‘find’ devices and preinstall applications and web links. There are however significant drawbacks to this methodology. Devices are essentially locked to the original computer that managed the setup, and require a back-to-base approach for management of devices needing the devices to be physically sent back to their original setup location. To provide many management features the devices essentially become locked down and any customisation, installation of other apps or data on the devices is unable to be backed-up or maintained by the user. This model would be suitable for applications such as creating class sets of devices that can be borrowed and returned, but never owned or customised. This kind of arrangement would be similar to a computer lab model.
- Devices can be set up to be loaned & shared to provide similar functionality as a computer lab
- This computer lab model is portable and less reliant on physical infrastructure.
- Provides a way of giving access to apps, interactive texts, rich media for classes and groups
- Manual process of setting up can be both difficult and time consuming. Additional equipment would be required for large deployments.
- Back-to-Base support requirements limit flexibility and increase the need for additional support and ongoing maintenance.
- Any devices that are loaned and are not using this model may violate Apple’s Terms and Conditions, which have specific requirements around a user’s Apple ID and how it can be applied to a device.
- There are options to reduce the management features on the devices but in these circumstances there is little benefit.
In a general sense a managed deployment would suit applications where devices are required to be managed, such as a lab environment, loan devices, or where support requirements need to be kept to a minimum and externalised.
Unmanaged rollout is essentially handing the device over as new in the factory default. This method requires a user to manually setup the device they receive. This is the model that the project deployed for all its trials in 2012, as it provided the most flexibility and allowed greater ownership and buy in from the students and staff. The users are allowed complete freedom to set up, install and use the devices however they want. The devices are essentially totally open to all customisation, installation of other apps and data on the devices is backed-up or maintained by the user. This model would be suitable for all other applications other than creating sets to be borrowed and returned. It is similar in scope to the Bring Your Own Device (BYOD) model, but takes advantage of the university for purchasing, logistical rollout and support. In terms of software, Apple has recently enabled the Volume Purchasing Plan, which would allow the university to buy apps in bulk and then distribute them to staff as redemption codes.
- Provides maximum flexibility and customisation from the user’s perspective to encourage personalisation and usage.
- Maintenance and backup is done via the user’s own machine so avoids the back-to-base requirement, especially applicable with distributed cohorts of staff and students.
- This method improves digital literacy by encouraging ownership of the device through responsibility. With the correct support this has seen the growth of significant and applicable skills and knowledge in staff and students.
- Requires users to set up the device themselves, which can be simplified by rolling out documentation, tutorials and face-to-face support.
- Users are responsible for backing up and maintaining the device. This is a simple process but does require users to learn the process.
An unmanaged approach does put the onus on the user to maintain and manage their own device. This goes against the traditional approach to provisioning technology at CSU and many institutions, however some of the distinct benefits to this are a greater sense of ownership and the development of better digital literacy. Users are required to learn and understand their device to a further degree, but are able to customise and personalise their experience to suit themselves. This model is a trade off between the provision of additional support and development instead of management services and infrastructure.
It is vital that part of the decision-making in this area includes support. Overall, the Project has seen a reduction in support requirements through the adoption of a single vendor approach, which has improved the efficiency and efficacy for our rollouts.
In the surveys, staff and students were asked to choose which support resources were the most useful from the following:
- Walkthrough Videos
- Interact Site
- Face to Face
Staff and students overwhelmingly found face-to-face support to be the most useful. This can be confirmed by the Project team, as it appeared to be the most productive way to rollout the technology, up skill staff, and provide feedback and ongoing support. During the staff trials the Project also tested the establishment of an organic community of practice. These events were informal but provided a chance to meet with technical staff. It was observed that these sessions would quickly concentrate on peer-to-peer learning, as staff shared their experience, so facilitation became the main role for the Project.
At the launch of the Project there was really only one tablet – the iPad. Throughout 2012 we have seen a huge range of android devices being released and newcomers, like Microsoft, come to the mobile party. The following table maps out the three main tablet candidates in the current crop of operating systems (May 2012) and how they compare in a number of key areas.
|Model||Apple iPad||Window Surface||Samsung Galaxy Note|
|Operating system||iOS||Windows 8 RT||Android 4.0|
|Device Price||$539 – $1009||$559 – $679||$585 – $835|
|Cover Price||$45 – $80||$140 – $150||$45 – $60|
|Total Price Range – RRP||$584 – $1089||$699 – $829||$630 – $895|
|Comments||* Pricing scale includes WiFi and 3G||* Pricing scale includes WiFi and 3G|
Moving forward towards developing a better plan for mobile will require collaboration and consultation between divisions, faculties, staff and students. The first step would be to establish a working group to investigate how CSU can support a rollout of mobile devices to both the staff and student bodies. This group will be tasked with providing details relating to the issues involved in a rollout, and a variety of provisioning models based on financial and sustainable practices.
This working party would be tasked with:
- Opening a specific and focused dialogue with each school and division to develop requirements, opportunities and affordances that mobile technology could deliver. Discuss the device specifications, software and hardware requirements for a range of discipline-specific tasks and possible research opportunities.
- Opening up dialogue with potential commercial partners and investors. Mobile technology could offer a range of new commercial opportunities for CSU, with many companies interested in establishing relationships in this area. This would be mutually beneficial in terms of exposure and marketing of CSU as a progressive and modern university and possibly lead to new income streams.
From the specifications developed the cost of provisioning models would be developed including ongoing management of devices, lifecycle and deployment. This should be done in conjunction with Finance and DIT to consider appropriate funding sources. Judging the current climate should be based on a 24-month turnaround of devices.