Many have at least one business process that can be solved with mobile technology, and these business processes are usually pretty unique to an industry - and in many cases, to a single company. With this in mind Pat Brans spoke to four CIOs about how their organisations were using mobile technology to better project manage their businesses.
Because there is so much variation in what mobile applications companies use, mobile solutions will never be mainstream, in the sense of being shrink-wrapped, and one-size-fits-all.
There will always be a high degree of customisation - much more than the amount of customisation required to fit an ERP or CRM system to a particular company's needs. But mobile technology is mainstream in the sense that most companies are using it in one form or another.
With this in mind, I asked the IT departments of some large organisations what exactly they are doing with mobile technology, what their plans are, and what pieces they think are missing from the marketplace. For this article I spoke with four CIOs:
- Rick Swartz: US Census Bureau
- Ray Vaughan: Royal Mail Group
- Robert Lee: Bibby Distribution
- Phil Pavitt: Transport for London
Mobility requirements
From a CIO's perspective there are two things that make mobility different from other kinds of IT services. Every mobile user has at least one of these characteristics - and many have both:
Users are outside company premises: Computing occurs in an environment that is not physically secure, and where IT staff cannot physically come over and help the user. This has several implications, including additional security needs and requirements to provide support and troubleshooting in non-traditional ways.
Users are moving around: You can't pinpoint a user - that is, his or her movement is virtually unlimited. Because the user is moving, he or she relies on a constrained computer, and either connects to the enterprise through a wireless network or works offline, synchronising with the enterprise from time to time. Security is an issue here as well: you need to protect against the possibility that the device is lost or stolen.
The problems - and the expertise required to solve the problems - are fundamentally different. Also the mobile community is generally more tolerant than the in-building community. Therefore, you have to build your SLAs, your support services, your messaging, your information flows, and your upgrade plans differently for two different types of users.
Now let's have a look at some of the company-specific needs.
US Census Bureau
The US Census occurs every ten years, the next one being in 2010. The Census Bureau sends a form to each household and asks the occupants to fill it out and send it back. Based on past experience, the Census Bureau expects that around thirty-five percent of the households will not respond, so they plan to send workers out to knock on the doors of those who don't respond and collect the data by asking questions. This process, called non-respondent follow up (NRFU), takes place over a period of around six weeks, and requires a half a million field workers, the vast majority of whom are hired just for the NRFU.
In order to perform a census, the bureau needs a good address list. In order to update the address list they had from the previous count, they perform a task called address canvassing, the next one of which will take place in early 2009, and will take around nine or ten months to carry out. Address canvassing requires over one hundred thousand workers who go out and physically account for new construction, torn down buildings, and any other changes to addresses.
Both NRFU and address canvassing are processes the Census Bureau wishes to improve through mobile technology. They have already developed a solution for address canvassing, and will use it in 2009; as for NRFU, they have scoped out a mobile solution, but they aren't yet sure they'll use it in 2010. The same customised devices will be used for both processes - the devices are normal PDAs but without some of the standard features, such as camera or email.
In both cases, security requirements include two-factor authentication, encryption, and device lock down and data wipe in the event the device is lost or stolen. GPS is also required in both processes in order to pinpoint households that might not have traditional addresses - a common occurrence in rural areas, where the postman might identify a house as something along the lines of 'the small house across from the old oak tree'.
Royal Mail Group
Royal Mail Group (RMG) have about thirty thousand managers, most of whom have some sort of mobile service, such as email on a PDA, or a laptop with a 3G card. They also use mobile technology to help in some key business processes.
This usually involves scanning items so they can be tracked: currently, they have seventeen thousand mobile devices to scan packages at collection points, and about seven thousand to scan packages within buildings. In both cases, data is synchronised through a fixed connection, when devices are placed in a cradle.
RMG is looking to gain competitive advantage by providing real time tracking all the way to the door. A step in this direction is something they are currently working towards: deploying twenty-five thousand devices in vehicles. The data will be synchronised in real time using GPRS, and devices will be docked in a cradle in the vehicle to ensure good battery life. Wi-Fi and fixed connections will be available at the home base, if for some reason GPRS fails.
Bibby Distribution
Bibby Distribution uses mobility in ways Robert Lee calls 'passé' - for example, mobile email, and PDAs and laptops with wireless adaptors for roaming users -, but the most important business processes they apply mobile technology to are those around their core business, which is making deliveries on behalf of clients. In particular, they rely on mobile computing technology when they undertake delivery during the night.
This involves unattended deliveries, since nobody is around at night. Delivery is sort of like an airlock: they put the merchandise in one side and it's secured, then the client takes it out the other side when ready. In this situation, service levels are absolutely extreme - that is, no failure is acceptable. They barcode the vehicles and each site has a barcode, so they have put in place a mechanism to prove delivery took place.
Bibby Distribution can automatically generate a message confirming delivery. Clients want this in different forms: some want a text message, some want email, and some want a good old fashioned phone call. Many clients only want to hear when there's a problem - for example, if a car is blocking the way of the delivery truck, or if the client has changed the lock and the driver can't open a gate. Being able to generate these alerts allows Bibby Distribution to overcome what would otherwise have been delivery failures.
Transport for London
Currently Transport for London (TfL) has around two thousand five hundred mobile IT users, where they define mobile users to include people with Blackberry-type devices, laptop users, and home workers. Within two years they plan to have a total of twelve thousand mobile users, with most of these users being operational staff. Phil Pavitt calls this their 'real mobility need': supporting staff who are out in the transport network - and need quick and accurate answers to customer questions.
If a customer asks a TfL employee when the next tube to Ealing is, the staff member will be able to use a mobile device to get that information right away and with complete confidence. At first, the information provided to operational staff will be limited to the mode of transportation they work for, for example, bus or underground, but increasingly, TfL will provide mode-independent information that will enable bus staff to provide advice on tube availability and vice-versa.
Beyond equipping operational staff with that kind of information, TfL would like to put the same information directly in the hands of customers. For this reason, TfL needs to provide a service that is as device independent as possible. This second phase, where customers get the information directly, will not occur until at least two years from now.
What's missing in the market place?
Each of these four organisations has different needs to be met by mobile technology.
Rick Swartz says that for them human interface is a big issue. A lot of people doing the NRFU are retired people - older people who are not computer savvy. And now they are going to be asked to fill out forms on handheld computers that are even more difficult to use than desktops. If the Census Bureau goes through with equipping workers with mobile technology for NRFU, they are considering holding a five-day training course for each of the five hundred thousand users. That's quite an expensive undertaking.
Ray Vaughan says that one of his biggest concerns is integration with back-end applications. He thinks they are generally using all the mobile technology in standard ways, but application integration is almost, by definition, different from one company to another. He is also concerned with the constraints imposed by limited battery life, limited memory, and unreliable network coverage. These three factors feed into each other, therefore, any improvements in battery life, memory capacity, or network services are welcome.
Ray also sees, as a problem, that the market for mobile software consists of big players on the one side, who don't get enough of their revenue from mobility, and therefore don't give it the attention they need, and small players on the other side, who have great software, but don't have the proper controls for release and configuration management. For a big rollout like the one RMG is expecting, you really need to have quality assurance and release controls to make all upgrades run smoothly.
Robert Lee told me they would like to implement an RFID solution, so that each vehicle has a unique tag, and the handhelds would serve as RFID readers.This eliminates line-of-sight requirements, and it allows them to more easily identify things at item level. But Robert doesn't think RFID is mature enough.
The technology is still expensive, and it's something the entire supply chain has to agree to use. For that to happen, either the supply chain has to agree on standards, or somebody is going to have to use readers that are flexible enough to read all the different kinds of tags used by the different manufacturers along the supply chain.
Phil Pavitt explained that TfL has lots of relevant information and their aim is to exploit this for the customers' benefit by enabling the traveling public to 'pull' information relevant to where they are.
However, the user interface is still weak and the re-purposing of information for a given device type remains an issue because currently some data is not fully readable. The early adopters can tolerate it, but the larger user population still has trouble with it. Phil summed up the problem: 'I'm waiting for a company to do what iPhone has done for phone users: provide a beautiful interface.'
Conclusions
Solutions designed for specific business processes are very much bespoke and the company-specific nature of business processes makes it hard to imagine that ever changing. The customisation required for these solutions brings about at least two challenges. First, there is the expense of building an application from scratch. Second, maintaining such an application means maintaining a staff of developers and support personnel knowledgeable in that unique application.
Security is very much a concern. The requirements are generally that there will be user authentication, and data encryption both on the device and over the air. Also it's necessary to be able to lockdown a device and wipe it clean of data when it is suspected that a device has been lost or stolen.
Likewise, support is a concern, but all the components required are already available. It's important to be able to correct problems in the field and avoid having the user make a trip to the office for repairs. For this reason, remote trouble shooting capability is needed; and in fact, this feature is already present in the market.
CIOs are still concerned with the human factor. Many mobile users aren't computer savvy in the first place, and mobile technology is inherently more difficult to use. The devices are smaller and the interface is constrained. On top of that, since business process-specific mobile solutions are generally bespoke, the applications are not common applications with a large population of users that can share learning experiences.
