A gamble on a relatively unknown technology four years ago is paying off for a logistics company, which is using the software to shave millions of dollars off its bandwidth connectivity costs. Today freight forwarding company JAS Global is leveraging a software-defined wide area network (SD-WAN) to run cloud applications, according to JAS CIO Mark Baker. Eventually, Baker hopes to use the SD-WAN as the backbone of a predictive analytics strategy to grow the business.
SD-WANs allow companies to set up and manage networking functionality, including VPNs, WAN optimization, VoIP and firewalls, using software to program traffic routing typically conducted by routers and switches. Just as virtualization software disrupted the server market, SD-WANs are shaking the networking equipment market.
Cisco Systems earlier this month agreed to acquire SD-WAN startup Viptela for $610 million in a bid to accelerate its market strategy. A big driver for Cisco’s move is that SD-WANs are gaining traction in corporations, with over 3,000 paying customers and 100,000 branches deployed worldwide, according to Gartner.
SD-WANs weren't on many CIOs' radars in 2012, which is when Baker first became acquainted with the technology. Baker was consolidating global IT operations, which after years of expansion had tacked on more subsidiaries, many of which ran their own accounting, warehousing and other transactional applications. After convincing the board and C-Suite that JAS should replace its custom software with available vendor applications, Baker selected one global ERP system from WiseTech Global, a logistics industry standard.
The big SD-WAN gamble
But there was a catch. WiseTech’s CargoWiseOne application ran off a single SQL Server database. Baker needed to find a way to ensure that the app could keep up with the logistics transactions required to support 4,000 employees spread around the world in 240 locations, including remote locales such as South Africa, Brazil and China. JAS was already supporting global apps and email with MPLS networks and VPNs but Baker was concerned about the potential costs of powering the crucial ERP software, which required sub-150 millisecond latency with a global MPLS network. Setting up and provisioning an MPLS system also takes several months.
WiseTech CEO Richard White recommended Baker take a look at SD-WAN technology from Aryaka. Baker took White’s advice and after looking at Cisco and Riverbed Technology, settled on the startup. Aryaka uses a global private network and SD-WAN software platform to deliver data and apps around the world with MPLS-grade connectivity and fidelity.
"For a global company it's easy to go from Atlanta to L.A. or London to Paris, but when you start talking about going across the pond or north and south hemispheres there is a huge latency challenge to overcome when you're lacking a traditional MPLS network," Baker says. "The challenge lies in delivering applications and services to business users and customers and optimizing that on a global scale."
Introducing the technology also proved challenging. Baker says the SD-WAN model "scared the hell out of" his nine Cisco network engineers because they didn't know how to operate it relative to traditional MPLS networks and VPN connections. "They didn't understand that it was not about managing the network at an ISO Cisco code level anymore," Baker says. "It was about working with application developers and business users and determining the optimum way to deliver an application anywhere in the world."
The big pay-off
Steering through the learning curves, JAS rolled out Aryaka in June 2012. The software connected JAS' ERP, email and other apps at a fraction of the cost of MPLS systems. Baker estimates he saved as much as 66 percent of bandwidth connectivity costs using Aryaka and cut provisioning time from six to eight months to less than a month. "It allowed us to deliver an ERP system on a global scale and overcome some massive local infrastructure challenges by leveraging simple internet bandwidth at a fraction of the cost of the MPLS services that we had or that would have had to pay for to put in place," Baker says.
Baker has begun to run cloud applications, such as the Zoom video conferencing application, on its SD-WAN network, replacing a hodgepodge of WebEx and Skype instances while boosting collaboration for JAS employees and trimming travel costs.
Baker is particularly excited about using the SD-WAN system for predictive analytics that will help JAS customers gauge whether their cargo will arrive on time. JAS plans to use Aryaka to ingest data from airline and ocean freight schedules and weather information and funnel it into a Microsoft Azure data lake.
Running machine learning algorithms against the data, the system will predict the probability of whether, for example, a Delta Airlines plane carrying a compressor part needed by workers on an oil rig will make its flight from Los Angeles to Houston. Baker says JAS customers are keen to "know the probability of an event happening" that would impact their shipments. Theoretically, such information could give JAS a leg up in negotiating customer deals.
"With the foundation of SD-WAN, we can talk about moving into machine learning and artificial intelligence, helping us weave JAS into the supply chain of each and every customer," Baker says. "[SD-WAN] has become a disruptive technology because it elevates network connectivity as a business value."