Showing posts with label Mobile device. Show all posts
Showing posts with label Mobile device. Show all posts

Tuesday, May 12, 2015

Mobile Device End of Life

Mostly when people think of enterprise mobile projects what springs to mind is the process of designing, building, and testing, a fancy new mobile application. When this new application goes live suddenly life is easier, the company makes more money, customers and employees are happier. (For a more comprehensive view read the benefits of enterprise mobility). So where's the "but" you ask? I'm sure you have guessed that not all enterprise mobility projects follow this same pattern. Equally as important as a new application is ensuring that an existing mobile business process continues to operate effectively.

Many companies with a long history of mobile applications have invested next-generation improvements to leverage the benefits of technological advances. Typically aspects of the end to end mobile solution may reach end of life. For example support may become less effective, skills in niche technology become scarce, devices and operating systems change. Additionally user expectations increase along with consumer technology.

In the traditional enterprise mobile space many organisations opted for rugged or military specification devices (and plenty still do). One of the benefits of these devices is that they tend to be supported by their hardware manufacturer for many years. Eventually though this many years does run out. At that point there is usually a scramble to think about new devices and then the doors open to options, alternate vendors, and solutions.

What should you do if you are tasked with taking on an end of device life scenario? Before we get started let's set the scene a little more. Consider Company X has a mobile solution that includes integration to a business system, offline capabilities along with some peripheral integration such as printing and/or scanning, and is leveraging a rugged mobile device that will shortly become unavailable and out of support. This maybe somewhat more complex than users of web applications moving from iPhone 6 to iPhone 7. However in concept the steps are similar, just greatly accelerated in the simpler example.

So now we have some context let's firstly consider the three "environments" of relevance. The legacy, the current, and the future. With all three it's good practice to do due diligence to ensure that you have cost effectively mitigated the risks.

For the existing legacy solution a great starting place is the solution documentation. Anything you can dig up on the existing business case, processes, technology stack, products, application, devices, support and integration will help prepare you for more detailed discussions. Consider for example that the standard operating environment or build of this kind of solution can be quite complex. Prepare a simple one or two page summary of your findings. Once you have the background it's a good time to talk with the stakeholders, users, and support staff to understand the requirements. Again ensure to take notes and prepare a summary.

Next the current environment phase is about understanding what is available in the market and how this will fit with the existing solution. It's time to begin the device evaluation process. For full details of this process check out the following article that covers how to select a device for enterprise mobility. A good process is to shortlist devices and then bring the hardware manufacturers and/or demo devices to the stakeholders to show the relative pros and cons. Don't forget that in our scenario it's not just the mobile device that is important but associated peripherals like printers and scanners.

Once a subset or limited number of devices have been established then it's time to do some testing. Determine the appropriate level of testing by evaluating the delta between your existing and target devices. For a full run down check out this article Testing Enterprise Mobility. A key consideration is to ensure that the software build can be completed. This can be greatly impacted by operating system and hardware changes and may require cycling back with the software providers. Once technically working it's worth field testing with end-users to ensure that any negative and positive aspects of the new device are understood. Ensure that all the testing results have been documented.

Finally consider the future. Change is rapid in the digital space. Is the solution sustainable? Are aspects of the technology now past a point of no return? Is replacing the device the best method to continue to get value from the solution. Perhaps it's more cost effective to replace certain parts of (or the entire) solution based on new software and hardware solutions.  To wrap up prepare a summary of findings along with a recommendation based on the facts.

Friday, February 8, 2013

Testing Enterprise Mobility

One challenge for any IT project is to ensure that testing is performed efficiently and effectively. As with many things the goal is to try to find a happy balance. In this case ensuring quality without spending forever on the testing cycle. Before I go any further my intention here is not to dwell on V charts, and testing frameworks but rather to talk specifically about testing enterprise mobile applications.

Key Considerations
  • GUI
  • Screen flow/work-flows
  • Integration
  • Mobile device/s
  • Communications
  • Infrastructure
  • Security
  • Roll-out
GUI, Screens, & Integration
Certainly a lot of the focus when testing enterprise mobility should be the application, its screens, and integration. Involving end users early (when managed correctly) can really help to ensure that the screen flows work under real use scenarios. Often functions can be tested independently during the development cycle. Device Emulators/simulators on a desktop can help rapidly test while waiting for the infrastructure and comms to be put in place. If the mobile application works in an offline way then ensure to test for data conflicts, locks, and test the resolution process. How can the solution recover from an integration error?

Mobile Devices
An important consideration that is sometimes overlooked is the wide variety of Mobile devices, models, operating systems, screen sizes, peripherals, and input methods that are being used in the solution. Don't assume that everything that works on Model X will automatically work on Model Y.

Battery life - does your shiny new mobile application drain the battery flat in minutes? Is it using lots of CPU when decompressing or encrypting data? Does it constantly require communications or gps? Schedule some tests to simulate different levels of usage.

Communications
Communications (or lack there of) should be checked. What happens when halfway through a transaction the comms is lost. What happens if comms is slow or poor? I've seen some creative ways to test this. For example underground car parks, driving through dead spots, or using a Faraday chamber. When projects first move from simulators to devices they are often cradled or connected via wifi. Ensure at some point to factor in tests based on real world communication scenarios.

Infrastructure
With infrastructure it's often not possible to directly replicate the production environment. However it's definitely worth looking at simulating traffic. Try running functional tests while ramping up the load and find out at what point problems occur. MEAPs often come with tools for capturing, simulating, and amplifying traffic.

Authentication & Security
Its obviously important to test the authentication and security mechanisms. Consider carefully how the authentication works and schedule appropriate testing. What happens when passwords expire or the user enters it wrong too many times. Some companies hire in specialists to attempt to compromise the system.

Roll-out
As part of the overall strategy consider using a technical go-live where the mobile application is put into production and tested without being used for real business. This may require assistance from the business to run through some transactions and then back them out. Once the technical solution is confirmed then one or two key users can be introduced to the solution. At this stage monitor the system thoroughly and adapt any changes as necessary. Once proven expand the user base gradually and ensure that performance is managed. On some projects it's necessary to initially throttle the performance so that the first user group does not feel negatively impacted once the entire community is live.

Please let me know your thoughts and any additional lessons learnt from testing enterprise mobile applications...