The consensus of panelists convened for the 2005 VoIP: Telephony conference here yesterday was that voice over IP eventually will transplant time division multiplexing as the technology platform of choice among telecommunications service providers, but not before a plethora of network and operational hurdles are cleared at both the service provider and customer levels.

“The world is going to change, and VoIP is going to be a part of it, but it’s not happening as fast as the hype,” said Mark Kaish, vice president of next-generation services for BellSouth, who delivered one of the morning’s keynote addresses.

A hindering factor is that potential VoIP customers often take a “if it isn’t broken, don’t fix it” approach to technology upgrades. “They recognize VoIP is the future, but they’re not in any hurry,” he said.

Another hindrance is cost, which can range from $600 to $1000 to upgrade a station in the enterprise to IP, according to Kaish. “This is a very tough business case unless you can prove that productivity gains” will result from the implementation, he said. In addition, it’s often difficult to merge voice and data staffs within an enterprise, most companies lack the security expertise to support VoIP and quality of service, and bandwidth concerns persist.

Also, troubleshooting is far less effective in a VoIP network than it is in a TDM network, said Mark Pugerude, senior vice president of marketing and business development for Broadwing Communications, who delivered the other morning keynote address. “There are lots of tools in TDM networks to fix problems, but not so in IP,” he said. “Typically you would triage to find problems but you can’t do that in VoIP because there aren’t enough tools.”

Nevertheless, VoIP is thriving in greenfield situations, where such deployments recently exceeded TDM deployments for the first time. “That will continue,” Kaish said.

Much of the discussion during the conference centered on the role of IP multimedia subsystem architectures in future VoIP networks. Frederick Lax, CEO of next-generation signaling and switching solutions provider Tekelec, characterized IMS as a “pretty complex” architecture that is a work in progress, and said no vendor currently has a complete IMS solution. He also said IMS has been over-hyped and has “received a pass on a lot of important issues that aren’t going away.”

“IMS isn’t a magic bullet but it holds enormous promise. Some solutions are ready now, and five years from now it will be a done deal. Service providers need to create evolution plans to make sure they’re around” when it happens, Lax said.

Phil Holmes, chief technical officer of BT Exact, in a later breakout session mildly disagreed with the notion that IMS has been over-hyped. “IMS is a relatively mature technology, but the applications aren’t as mature,” he said.

Roger Heinz, vice president of Lucent Technologies’ Convergence Solutions Group, agreed that applications represent a tipping point. “The key--for both vendors and service providers--will be whether we can deliver applications that end users want,” he said. “That will determine whether IMS will be successful.”