As if selecting, leveraging and managing information technology weren't enough for fleet owners to deal with, now they've got to get a grip on the whole idea of “cloud computing.” And then there's another wave of innovation ahead to absorb: business intelligence (BI).
The downside of cloud computing is that it's one of those pesky paradigm shifts that force all IT users to look anew at how they run their operations and at which IT tools they use. But the upside of this new method of storing, retrieving and acting on data far outstrips running through those initial mental gymnastics.
Cloud computing, despite its whimsical and wispy moniker, makes eminent sense once it is fully laid out. What's more, it is the way station, if you will, to making fully data-driven “business intelligence” — another rising IT buzzword — possible to implement, even at the fleet level.
The simplest definition of cloud computing is that it is — drumroll please — Internet-based computing. One could argue that “grid computing” would be a more accurate description as what is generally described as the “cloud” (or Internet-hosted capability) seems to directly parallel what we all know and love as the electrical grid.
At any rate, cloud computing is what the IT professionals call it, so let's get on with what's in the cloud, why it would be placed there and, above all, why should a fleet owner care. Cloud computing is really just a shorthand description of a new delivery model for IT services that bypasses the need for a client-server or mainframe system. Instead, the Internet is leveraged for remote computing.
Key advantages to this remote approach include lower system startup costs for the service provider and the end user as well as a much greater degree of scalability — the ability to grow the system quickly as needed — than with traditional software application delivery methods that depend on hardware owned or leased by the end user.
Cloud computing is also compatible with yet another popular IT expression: data mining — extracting nuggets of data that may have been compiled originally for one specific purpose and then looking at it from different perspectives to better inform tactical and even strategic business decisions. For example, pulling out dispatch data can enable a fleet to determine which of its customers are generating the most revenue and which ones are actually costing money to serve.
“Cloud computing as a term within the IT world has been around for awhile,” points out Stephen Pembridge, senior business intelligence analyst with TMW Systems. “It means to go outside your company's own IT space and rent or access by subscription software-based solutions. The operating system will reside within the cloud and the resources — the services — the customer needs can be ‘brought down’ to their workplace devices as needed.
“The idea,” he continues, “is simply that as the Internet becomes more expansive, there are more resources a user can access, again by reaching out of their own sphere and into a cloud controlled by a service provider.”
Another important aspect of cloud computing is that it does not have to be implemented as an either/or proposition. Moving new or even existing systems to run “in the cloud” need not be done wholesale or overnight, but at a pace and in a manner determined by the individual fleet buyer.
Indeed, there are fleets that may already be engaged in cloud computing without even knowing it. Some experts contend that cloud computing should properly be viewed as multi-levels — not unlike real clouds — starting from a basic level that trucking software users will most likely encounter on up to more loftier levels.
THREE CLOUDS
That basic level is analogous with the Software-as-a-Service (SaaS) delivery method that many fleets are already aware of, if not taking advantage of, courtesy of their software suppliers. SaaS enables fleets needing trucking management software to receive it at the lowest monthly cost possible and without having to host or control the computing power that drives it.
Bear in mind, no matter what the service is called, the cloud-computing provider will be delivering its applications via online links accessed from the web, usually through nothing more exotic than a web browser. The actual software and the data it generates will be stored on servers and remain under the control of the application provider, which means the fleet need not dedicate IT staff or material resources to maintaining the application.
SaaS is essentially software on demand. The given SaaS provider licenses an application to its customers via a subscription “pay-as-you-go” model, which in trucking is typically paid as a monthly fee. SaaS lowers the cost to access software because it does not require the user to license the use of the application on each device that will run it. And when delivered via this type of cloud computing, software updates can be done rapidly and automatically.
“The industry will see a gradual ‘creep’ toward wider use of SaaS, especially smaller trucking operations that need to ramp up with software quickly,” points out Mark Cubine, vice president of marketing for McLeod Software. “For example, our freight management product iLens is configured as a pure SaaS product and because of that, we can get a customer going live in weeks vs. months.”
However, Cubine cautions not to think everyone is headed for the clouds just yet. “Certainly, third-party logistics providers and freight brokers are in the forefront of adopting SaaS as they are leading-edge types in general and typically do not have a great deal of IT infrastructure already in place. But I would not say at this point that we are on the verge of a headlong rush to SaaS in this industry.”
Indeed, while the next logical thought would be to expect that fleet users can access the innumerable SaaS clouds being set up by solution providers, the down-to-earth reality is that information technology is not yet that advanced.
“Right now,” points out Paresh Nagda, CTO for Navman Wireless, “there is no tool or technology readily available to let you go from cloud to cloud and to take data from each, and then use it in a meaningful way.”
Nagda explains that in his view, there are three types of clouds floating out there. The first, as described above, would be a SaaS cloud. “That's where a given application is sitting in the cloud and the app is maintained by the service provider with the customer using the app by linking with it in the cloud.”
He says the next highest cloud could be referred to as “Platform-as-a-Service,” in which both the software and an interface to build on it exists within the cloud. Nagda notes that salesforce.com is a good example of this approach.
The highest-level cloud in Nagda's view would be one holding infrastructure within it. “Here you would be talking about optimizing the structure to be able to use it on demand and also make it very easy to scale it up or down.”
McLeod's Cubine contends “there's a lot of confusion between cloud computing and Software-as-a-Service concepts. Clouds are genuinely where one runs their applications, and there are multiple programs on multiple computers in places I may not even know about that are serving all of it up to me. I may have a composite of a dozen different programs from a dozen different companies I use, and they may all be hosted in different places with data yet in different places. And there is the whole idea of a public cloud vs. a private cloud. A private cloud is where the whole cloud is contained within a single company and controlled and supported by that single company.”
DATA MINING
SaaS refers to where I don't have the server or the software on site, I don't own either, but I buy it on a subscription basis from somebody who is hosting it all for me,” he continues. “Then there are ‘hosting models’ where only the hardware is [in the cloud],” such as may be obtained by renting from such operations as Rack Space or even Amazon. If I choose to use those hosting models, I am deciding to quit owning hardware and to rent it instead. But I still buy software and have it installed on that rented hardware someplace so I can run it.”
Cubine points out that with SasS, a company's data is separated and not shared with any other firms. “SaaS is the web-services thing,” he adds. “I buy a service in a SaaS model and integrate it with my other programs. My programs then use a secure web connection to talk back and forth with that web service from some other place; an example would be doing mileage look-ups from ALK's PC Miler program.”
Navman's Nagda says data mining can be done within an SaaS cloud, but it “must be an integrative offering; that is, the [software] provider must provide tools for customers to mine their data. This is already taking effect in the marketplace. In this case, one company, such as ours, would do the [data mining] implementation and all our customers could access [that feature].
“In our case,” Nagda continues, “we will collect and store data for as long as each customer wants us to. We then offer them reporting capabilities and they can look at their data for a particular period to review and act on key performance indicators, such as jobs completed, average miles for day, week or year, etc. All this allows them to develop benchmarks and to see where any problem areas are. They can eventually ‘build out’ [the mining] as well from there, such as to the regional level.”
Nagda contends the advantages are clear to users: “There are lower costs and more streamlined operations for the customers. They do not even need to get their IT [department] involved in this.”
SLICE AND DICE
He recommends that fleet managers who want to get in on these advantages ask their solution providers if they have data storage and the tools for mining data within their hosting cloud. “Also ask how granular the data will be and what dimensions will be available to the user when mining,” adds Nagda.
TMW's Pembridge says the SaaS cloud “lets the consumer use the service(s) as needed by taking software code or data ‘down’ from the provider. There are functional advantages to this, including benchmarking enhanced by drawing from a wider anonymous pool of data [from other users]. The cost savings alone are significant. A user can get the benefit of new functionality without making a huge investment [in software] up front. SaaS essentially allows you to ‘rent’ the server space and also provides much greater flexibility for updates, etc.”
Don George, TMW's manager of platform independent development, says the “next part” of cloud computing is to move toward greater use of what he terms “true data warehousing” so that information stored can be more thoroughly “sliced and diced” to mine value from it. “The idea is to deliver exactly the information the customer wants,” he states, by putting the data into a structure that can be easily accessed. In our case,” he adds, “we have a wide variety of customers and they have different needs.”
Pembridge points out that TMW launched its Data Warehouse tool last year as part of its ResultsNow system. He says it makes it easier to perform and improve data mining and statistical analysis. “It all leads to lowering infrastructure costs, then more effort can be put into actually using the data to make business decisions.”
Not wanting anyone to miss the sky for the clouds, McLeod's Cubine says the biggest news in fleet computing will be what is already taking flight across the wider IT landscape: business intelligence. BI refers to using computer-based tools to identify, mine (or extract, if you prefer) and analyze business data, such as sales revenue by load or customer or determining costs vs. revenues. BI typically provides historical, current, and predictive views of business operations and includes not just data mining but also such tools as benchmarking and predictive analysis.
“Data mining has come to be called business intelligence,” Cubine explains. He points out that BI is not dependent on any of the [cloud] architectural concepts for IT systems. “BI is the concept that I am going to get data from many sources in my systems, put all that data into context, and finally serve it up to users. They will then be able to make decisions in a way that brings exceptions or critical ‘call to action’ data to the forefront, along with the analysis needed to make the best decisions and/or to explore the root cause relationships around a problem.”
CUBING IT
I think BI is about 20 times more important than the whole SaaS/cloud subject of today, because it is an area where companies are actually finding ways to generate competitive advantage and real business improvements right away,” Cubine continues.
He says McLeod plans to roll out a business intelligence module in a few months. “The idea behind it is to move into an environment that will pull multiple sources of data and place it into a separate database where the aggregated data can be used to perform complex analyses. Data tools called ‘cubes’ [developed by Microsoft and others] will then be used to attain a three-dimensional context so more pertinent, actionable data can be obtained and then sliced and diced to enable, for example, easily seeing profitability by route, customer, etc. It will also give us the ability to take a massive amount of data and by applying context to it fully analyze it for trends to ultimately make more informed business decisions.
“I wish I had a crystal ball to know how far the whole world of IT is going to march down this path toward SaaS and other cloud-hosted solutions,” Cubine adds. “The market is going to decide. For our part, we built our products with a technology that easily allows us to slide in any direction the market and our customers want to go.”
Keep your eyes open and your head up, because at the rate new innovations are coming at trucking, cloud computing and business intelligence are not likely to remain the newest and greatest things in IT for long.
