Project Santa: Developing mobile apps for multiple device types

At SAP TechEd Demo Jam two months ago we showed how we plan to make our consultants lives easier with mobile technology. The morning after demo jam I had a proper headache from celebrating our second place as well as an email inbox full of emails all asking: “When can I have the app?”

In the meantime we’ve kicked of Project Santa, which is all about driving efficiencies in the way we work. While the demo jam entry really was about floating some ideas and giving an impression of what a solution might look like, Project Santa makes those and other ideas reality for Bluefin. We’re going to use SAP’s planning tools for our resourcing process and a mobile application to make it easier to record time against projects.

In this blog series I’ll discuss some of the key challenges of the mobile application part of Project Santa and how we’re addressing them in our specific situation – clearly without any intention to provide universal or comprehensive answers.

Developing mobile apps for multiple mobile devices types

So let’s dive right into our first challenge, which is that we are faced with a wild mix of mobile devices. In a recent poll we found that our workforce uses pretty much any mobile device type out there: Android, BlackBerry, iPhone, iPad, Palm, Symbian, Windows Mobile, Windows Phone, etc. So how do you go about creating mobile apps for such a long list of target devices and be ready to support whatever comes next?

Catering for all of them seemed pretty daunting, but when we looked at the numbers in the poll we found that if we focus on just iOS and Android devices we can actually reach around 85% of our smartphone equipped workforce. That’s not bad for a start, so we decided to concentrate on those two device operating systems for now but with an eye on how we can expand the reach further.

Mobile platform or micro app?

I thought hard about using a mobile application platform like SAP/Sybase Unwired Platform or Syclo Agentry. We partner with both vendors, so have the knowhow in-house to build on their platforms. In my opinion however today’s mobile application platforms really shine when there is a requirement to build mission critical mobile applications that need to work without network coverage. This is understandable given that this is where the money was in enterprise mobility over the last ten years. I’d be keen though to see a more holistic approach to mobility that addresses the needs of small and complex scenarios for both internal and external users on a wide range of target devices. The recent messaging from SAP and other mobility vendors seems encouraging and from what I’ve seen so far they’ll make some major steps forward in 2011.

There’s nothing like the present though. The “lighter” approach available today is to write a mobile application that directly connects to the backend system whenever it needs to retrieve or update data. This style of applications is commonly called “mobile enterprise micro apps”, while SAP seems to prefer the term “mobile instant value applications”. The overall complexity of this paradigm is much lower and security doesn’t tend to be such a big issue as no data is stored on the device. The consequence obviously is that the application won’t work if there is no network coverage.

Our mobile timesheet can clearly be classified as a mobile enterprise micro app, but that still leaves us with the question of how we’re going to make it work across a number of different device types?

Cross-platform compatibility with web technologies

During the run-up to demo jam I also experimented with using web technologies wrapped up into a thin native mobile application to achieve cross platform compatibility. But let me explain this in a bit more detail…

All modern smartphone SDKs can embed HTML content into applications via a control. The idea is that a thin native app only has to start up the HTML pages that are also on the device in fullscreen mode, so that application development effectively comes down to writing HTML, CSS and Javascript. All that is then needed to port it to another platform is to re-write this thin native wrapper and package it all up again. Luckily the guys at PhoneGap have done much of the legwork already and also added access to many typical mobile device services like the camera, compass, geolocation, etc. In addition I’ve added jQuery to the mix for AJAX access to backend services and the recently released alpha version of jQuerymobile for a consistent user experience and to deal with the subtle differences between browser implementations.

The initial experiments were promising so we got together for a joint proof of concept coding session. The developers involved grasped the ideas of the framework instantly with almost no learning curve thanks to the familiar web technologies. Within a day we had a rudimentary prototype up and running – both on iPhone and Android – and three days after project kick-off a screen prototype was presented to the to the business owner. The project team is now heads down in development mode so I hope we can share some lessons learned in a couple of weeks.

Other challenges…

In my next blogs I’ll look at other challenges (e.g. mobile application deployment, user adoption, etc.) that we’re tackling as part of Project Santa and any lessons learned. Please do let me know if there is a certain aspect that you’d be keen to read about.