Three guys walk into a headend... and decide to make careers out of cable
Charlie Elliott never got his engineering degree. He fell in love, he says, got married, and needed to get a job. The one he got was building electric harnesses for a company that made coin-change slot machines.
It turned out Elliott’s boss’ husband worked for a cable TV company, and his boss recommended him for a job. Elliott joined Ottawa Cablevision in 1972. He ended up working for four cable companies without ever changing jobs.
Cablevision was purchased by Selkirk, which was purchased in 1988 by McLean Hunter, which was purchased in 1994 by Rogers Communications, and that’s how Elliott ended up at Rogers.
|The 2008 Emerging Technologies award winners|
The SCTE bestows three awards for engineering excellence every year at its Conference on Emerging Technologies: The Polaris, Star of Integrity, and IP Innovator Awards.
The Polaris Award is given to an engineering manager who has displayed exceptional achievement and commitment to the development and/or deployment of hybrid fiber/coaxial (HFC) networks. Nominees should be engineering managers at the system, MSO or vendor level. Nominees must be members of the SCTE to be considered. The 2008 Polaris Award was sponsored by CommScope.
The Star of Integrity Award is bestowed on a member of the SCTE who, as an engineer or manager, has contributed to the efficiency and effectiveness of next-generation networks in delivering on-demand services. Nominees must be members of the SCTE to be considered. The 2008 Star of Integrity Award was sponsored by Arris (C-Cor).
The IP Innovator Award is given to an engineer or manager who has displayed innovative achievement and creativity in the advancement and development of IP networking. Nominees must be members of the SCTE to be considered. The 2008 IP Innovator Award was sponsored by Juniper Networks.
The SCTE will begin accepting nominations for the 2009 Awards later this year, but can be contacted at any time about the nomination process.
The Polaris Award recognizes engineering managers, and Elliott has a disposition for it – he enjoys finding solutions to problems that go beyond just the equations.
“You have to deal with people and processes. You mix them up in a bowl to get to a solution. That’s the fun part,” he said.
One of the solutions he’s most proud of was among the first he was involved in. Selkirk had introduced a home security service in its systems in Florida. When Selkirk bought Ottawa Cablevision, it wanted to duplicate the service in Ottawa. The problem was that the home security application was predicated on having a two-way plant, and at the time, Ottawa was one-way. Elliott put together a team, which figured out a way to activate the return path in six months.
Elliott became general manager of the company’s system in Toronto. He got to roll out home security in Toronto, too.
Experience upgrading plants to two-way would come in handy later, in 1996, when Rogers needed someone to manage the rebuild of all of its systems to make them two-way for the launch of Internet services.
Rogers has been particularly successful propagating new technology. The trick to that, Elliott explained, is to create a set of tools that can be used, then take the time to educate the entire company about how to use them.
One promotion followed another until, as vice president of national network implementation, he was given responsibility for three major rebuild projects upgrading the access network from 450 MHz or 550 MHz one-way systems to 860 MHz two-way systems.
He also presided over the development of a DWDM architecture in the access network to eliminate expensive fiber builds in remote areas.
Rogers also credits Elliott with managing the development of a unique deployment of fiber-to-the-curb architecture for green field deployments.
Elliott’s team, along with the company’s vendors, came up with a GigE optical link, using in-field optical multiplexing, which allowed nodes to be daisy-chained. Consequently, multiple nodes could be multiplexed, but as far as the hub was concerned, would still appear as a single node.
The low incremental cost of this architecture allowed Rogers to go 100 percent node plus zero (node + 0) in green field deployments, the first major MSO to do so, according to Rogers.
“It happened quick, and it saved us millions of dollars,” Elliott said.
It’s customary to acknowledge one’s engineering team, and Elliott does. His success, he said, is predicated on relying on others.
“I early on realized you can’t know everything,” Elliott said. “I rely on my guy to say whether something works. You don’t have to do everyone else’s job; you rely on them to do their jobs well. I’ve had the great fortune of always having people around me who warranted trust.
“You have to have confidence in the people around you to succeed – and indeed fail,” he continued.
“A bad decision is better than no decision,” Elliott said. “You can always correct a bad decision. But no decision leads to nothing. If you have no failures, you’ve failed. It means you haven’t taken enough risks. If you’re too safe, you’re missing things, you’ve lost opportunities,” Elliott said.
“I got that from my Dad. He told me that if you don’t make mistakes, you don’t learn anything… I’ve learned a lot.”
One of Elliott’s last projects is a fiber-to-the-home implementation, also for green field applications. It uses passive optical splitters and will have its first deployment in a new Ottawa subdivision this year.
He doesn’t expect to see Rogers’ FTTH deployed, however. Elliott is taking early retirement to spend a little more time with that girl he left college for – and some kids and grandchildren, too.
Growing up, Greg Griffis liked math and science, and was interested in engineering in general, but he’d decided he was going to be a doctor. At Georgia Tech, Griffis realized, however, that he most enjoyed electrical engineering. So he began thinking about biomedical technology. Georgia Tech had a co-op program, and through the program Griffis got to spend some time working at Nortel.
… And he was hooked. When he graduated, he went to work for Nortel. He started out with traditional TDM switching, but quickly gained experience with IP/MPLS (multiprotocol label switching) technology. The future was going to be IP, and Griffis was right in the middle of it.
And then Nortel shut down the division, laying off everyone, including Griffis.
Intent on staying in the Atlanta area, his choices were BellSouth or Cox. “And there was more going on at Cox,” he said. Certainly more that was IP-based; there was packet-based voice, PacketCable, video-over-IP....
Griffis is now senior network design engineer focusing on backbone design, responsible for figuring out an efficient and operationally easy way to align voice, video, and data products in the network.
“That’s the neat thing about my role – and the hard part, too,” Griffis said. “I have to focus on all the networking. I can’t focus on just voice, or just video, or just data.”
He earned the IP Innovator award not only by figuring out a way to use Cox’s packet switched network as a converged transport network for real-time voice, video, business VPN, and high-speed Internet, but also for creating a policy to implement it across Cox’s decentralized operations.
In other words, the solution goes beyond network equipment. IP is still new, especially for cable veterans who have been working with RF for 20 or 25 years. Griffis has been compelled to explain and demonstrate that IP-based technology can carry video and voice as reliably as analog technology. And because Cox is a decentralized organization, he has to do so repeatedly.
Each time Cox brings one of its systems into the network, Griffis’ team familiarizes itself with the specifics of the local network, and before talking with the local engineers, they come up with different possible implementations.
“That gives them the opportunity to figure out their options,” Griffis said. “We aren’t just going out and replacing things. Local autonomy gives local managers the chance to respond appropriately to local competitive threats. So we’re trying to strike a balance between a full-blown national rollout and local flexibility.”
That’s probably what Cox means when it lauds Griffis for his “ability to influence and lead initiatives that require organizational alignment from a variety of operations within Cox…he simply understands the big picture and how to move a team from the beginning to the end more effectively.”
Cox also credits Griffis with devising a creative approach toward leveraging its existing IP/MPLS infrastructure to save money on a parallel leased frame-relay network that was draining capital.
Cox Communications has 21 primary locations within its enterprise network. Prior to its service provider backbone build-out in late 2001, these sites were connected by a third-party frame relay network. The VPN was then migrated to the Cox backbone using BGP/MPLS VPNs between each of the sites at a tremendous cost savings to the company.
The company also had to include remote Cox sites that were not a part of the iBGP mesh. Griffis came up with the idea of combining L2 Pseudowire technology to concatenate these sites into the BGP/MPLS L3VPN. That allowed Cox to collapse the entire frame-based VPN onto the backbone. This VPN not only carries all internal VoIP and data between markets, but also provides connectivity for Cox’s Customer Call Center Virtualization initiative.
The ongoing challenges keep Griffis jazzed about working for Cox. “Every year there’s something different,” he said. “Last year it was business services and national VPNs; now it’s HD video delivery across the nation.
“It’s constant learning,” he said. “If you don’t, you get obsolete pretty fast.”
STAR OF INTEGRITY
Like many engineers, for Esteban Sandino, it started out with a facility for math. Once he got to McMaster University, though, Sandino had a hard time choosing among the many different disciplines that captured his interest.
Should he pursue something purely academic, or something applied? The latter, he decided, something that would allow him to apply his skills. But what? Nuclear physics (McMaster has a reactor)? Biomedical?
“Cable was not my first choice,” Sandino said. “I thought I was going to get into semiconductors. It was purely by chance I got into cable. I answered an ad and it was Rogers.”
Sandino says he was fortunate to start out at Rogers, because of the company’s penchant for volunteering to try new technology. “Somebody comes in with a box and Rogers raises its hand,” he said.
While there he was involved with one of the first implementations of digital modulation in cable (“We had a full rack of equipment for 64 QAMs”), and one of the first tests of cable modems from LANcity.
“I just liked it so much, I just stayed there,” he said. “Sometimes there’s what you plan – what you mapped out – and then there’s what you end up doing.”
From Rogers, Sandino moved to TCI Communications, which became AT&T Broadband. He did a stint at Aurora Networks and is now director of engineering in Comcast’s Mountain Division.
What’s kept him in cable – like Elliott, like Griffis – is that there are always new challenges. “That’s the thing about the cable industry; just when you think the technology is stable, something else comes along.”
At TCI, Sandino was a principle engineer working on one of the cable industry’s developments of DWDM optical networks, helping to test early DWDM products from Arris, General Instrument (Motorola), and Harmonic, among others.
Later, he served as chair of the SCTE HMS standards committee, helping to develop and implement the hybrid management sub-layer (HMS) standard.
Sandino was given the Star of Integrity Award for his recent work as Comcast’s lead engineer responsible for designing and implementing on-demand networks in seven major metro areas in Comcast’s West Division. That division serves over 5.5 million of Comcast’s subscribers.
Comcast’s approach was not to just install VOD servers, however; VOD is just one element of a number of interconnected projects, and Sandino has responsibility for many of them. He managed the construction of IP video transport networks and the storage, edge and management networks that support over 100 million on-demand streams a month delivered to customers this year.
All of that is now largely accomplished, and Sandino is now the principle engineer responsible for the trial and implementation of switched digital video, using new IP multicasting techniques, new session resource manager (SRM) and edge resource manager (ERM) interface standards and techniques, and new electronic program guide (EPG) interfaces. He also is directing Comcast’s efforts to test “next-generation” on-demand using the new SRM and ERM techniques.
“I know of no other industry as flexible, as quick on its feet, as this industry – and that’s not just the MSOs, that’s the vendor community, too,” Sandino said. “Every year there’s something new. What we’ve accomplished over the last 10 years has been amazing.”