Ask the Experts: IP-Addressable Video Technology

The most dynamic area of growth in the security industry over the past decade has been CCTV and surveillance. Not only has the double-digit revenue driven this market to unparalleled heights, but the CCTV technology available has evolved beyond the devices imaginable five short years ago. Industry sales vaulted from analog technology to digital almost overnight, and has since migrated to IP-addressable and Web-based systems just as quickly. All this growth has left many users and integrators taking cover as one technology explosion after another changes the face of the CCTV world they once knew.

Participant Biographies
Henrik Friborg is vice president of partner relations and co-founder of Milestone. He has been working in the IT industry for 20 years and is now using his insight working with IP video surveillance and security-related products. Prior to founding Milestone, Mr. Friborg was head of development in a software company developing mission-critical real-time computer systems for the banking industry.

Eli Gorovici, president and CEO of DVTel, has more than 10 years of senior management experience in the digital data communications industry. Previously, Mr. Gorovici was vice president of global sales and marketing for NICE's Visual Interaction Management Division. His management of the introduction of the NiceVision DVR to new markets established NICE as one of the top players in the digital video recording market.

Pete Lockhart is vice president of technology for Anixter Inc. In this role, he is responsible for research related to new product analysis for cabling and connectivity products for copper, fiber, wireless and associated communications solutions. Since joining Anixter in April 1988, Mr. Lockhart has been the director of technology and marketing, copper solutions, as well as the marketing manager and product manager for copper network cables and data cables.

Ian Ehrenberg, vice president and general manager of NICE Systems, is an expert in the emerging trend of smart technologies for video surveillance security applications. Prior to joining NICE Systems four years ago, Mr. Ehrenberg was involved in developing smart technologies for the telecommunications and graphics industries.

Ed Telders has been security manager for PEMCO Financial Services for 16 years. His responsibilities include physical security, information security, corporate contingency planning and safety programs. Mr. Telders has been providing security management for information and physical security in the banking, insurance and financial industries since 1981, from Fortune 500 to medium-sized companies. He is a contributing technical editor for ST&D.

As the general manager of Axis Communications, Fredrik Nilsson oversees the company's operations in North America. In this role, he manages all aspects of the business, including sales, marketing, business expansion and finance. Mr. Nilsson previously served as Axis' director of business development, leading the U.S. team responsible for growing the company's network video sales in various sectors. He has also served as the director of sales and marketing.

Don Taylor has served as vice president of marketing for Dedicated Micros Inc. for more than two years, in which time he has expanded the company's product offerings and market focus. He previously served as director of marketing for Sensormatic, and held several marketing director positions at Monarch Marketing Systems in Dayton, OH.

Ariel Silverstone is the chief information security officer at Temple University. He has more than 17 years of experience and specializes in consulting on the implementation of management information systems and networking systems. Mr. Silverstone's certifications include CISSP, CBCP, MCSE+I, MCT and CCNA. He has contributed to, authored and reviewed more than 30 published books.

Ted Brahms is a technical support specialist for Sanyo Security Products. He has more than two years of experience supporting CCTV products, five years' experience as network administrator and desktop support specialist, and one year previous IP video experience with early-generation IP CCTV cameras used for Webcasting applications and preschool classroom monitoring.

ST&D: Is IP-addressable video technology available and proven?

Ed Telders, security manager, PEMCO Financial Services: Yes, indeed. Our installation now uses exclusively IP-addressable video technology.

Pete Lockhart, vice president of technology, Anixter: The answer is a very strong yes. Although the early IP/ethernet-based products do have a number of operational issues, the next wave of technology with the new A/D and video processing and compression chips have greatly resolved most of these.

Eli Gorovici, president/CEO, DVTel: Our more than 200 installations with over 10,000 cameras is solid evidence that this is a proven technology across all market segments. The growth of this technology is quite impressive: If we look just at network video installations, the category has grown from around $10 million in 2002 to over $200 million this year?and we're still in the early adoption phase.

Henrik Friborg, vice president of Partner Relations/co-founder, Milestone: Yes, the technology is a reality today. The fact that Milestone has sold approximately 7,000 installations and that many of our customers are returning customers proves this. Our installations range from smaller installations with a few cameras to larger, mission-critical installations with several hundred cameras.

ST&D: Is IP/ethernet network infrastructure available?

Don Taylor, vice president of marketing, Dedicated Micros: We find the infrastructure more available in larger corporate facilities and environments. However, any size company can implement an IP network, even with a single recorder.

Ted Brahms, technical support specialist, Sanyo Security Products: In the United States most businesses have PC/data ethernet networks in-office. Many businesses also have or can acquire broadband access to the Internet via DSL or cable technologies. Current infrastructure is there in most urban areas, but suffers from lack of 100% uptime reliability, varying connection quality and bandwidth availability, and competition for bandwidth on LANs from other new applications such as IP telephony. Infrastructure reliability will have to improve for over-the-Internet remote applications, and bandwidth usage will also have to improve for IP video to be accepted as a dependable technology in the security industry.

Friborg: Yes, this technology is certainly very mature, since it has been the backbone of data networks and the Internet for decades. Intensive competition has pushed technological advances forward at a very fast pace and has pushed the prices down. Today you can buy a 1GB network at a reasonable price. Next year price/performance will be even better.

Lockhart: Most enterprises today have very robust ethernet infrastructures with multiple redundancies and backups and with strong cyber-security designs. The issue is more to the design of parallel networks or integration of IP video onto existing infrastructures using available layer 3 switching and QOS strategies. Today's enterprise customers' ethernet-based networks are ready for IP Video.

ST&D: When do you estimate that 50 percent of the cameras being sold will be IP-addressable, and what are the major obstacles to overcome in order for that to happen?

Taylor: IP cameras offer several advantages for some specific security applications, but the advantages are offset by several disadvantages that will prevent them from reaching 50 percent of deployed cameras in security applications for several years. The primary issues that will slow their widespread deployment in security applications include the cost of the cameras, high network bandwidth consumption and the fact that every IP camera in a security system requires a dedicated IP address. Another drawback is the lack of reliable video storage and archiving. While some IP cameras have a modest amount of built-in storage to archive video, they are vulnerable to extended IP network outages that will result in permanent loss of video. That being said, we certainly see the market trend moving towards this technology; thus, every DVMR we sell today is IP addressable.

Telders: I would venture my opinion that the conversion is inevitable. Many of the installations that are not using (IP cameras) today are doing so simply because they have a pre-existing hard-wired installation. The other limiting factor is lack of familiarity with IP-enabled equipment. Many potential users of this technology may have difficulty determining how to work effectively with the IT infrastructure to establish a new installation of this kind.

Fredrik Nilsson, general manager for Axis Communications: Today, we estimate that less than five percent of the cameras sold for security and surveillance applications are IP-addressable. This number is quickly growing, and analysts estimate that by the year 2008 more IP-addressable cameras will be sold than traditional analog. I think this is a correct estimate.

Friborg: The biggest obstacle for growth is the lack of IT and networking knowledge among the CCTV resellers and installers. Larger system integrators are now in the process of gaining the required in-house IT skills, but many of the smaller installers will not be able to do this, so they will be left behind over the next five to 10 years.

Lockhart: The hybrid solution of installing high-quality analog cameras connected to IP encoders must be figured into this equation. If both of these solutions are called "converting to IP camera technology," then three to five years is probably that crossover point. The biggest single obstacle to IP camera adoption is the VMS software required to turn large quantities of cameras into a seamless and workable network and how it will integrate into existing subsystems.

Brahms: I expect that for IP video cameras to reach 50 percent or better will require further product and infrastructure improvements. If security product manufacturers would meet and establish industry-wide standards for video compression and cross-vendor system compatibility, they could help facilitate the growth to 50 percent of sales. This would simplify integration of IP video equipment and applications from various manufacturers and assure end users of the future viability of IP video equipment purchases.

ST&D: Does IP-addressable CCTV offer clear-cut customer benefits? What are they?

Telders: Cost savings. The cost of establishing a new hard-wired solution is high due to skilled labor costs and maintenance of a new set of wires. Using existing network connectivity as an alternative can simply be more cost effective. I also see this trending toward a commoditized market, and as a result there will be price competition for market share and units will become less expensive over time. Network maintenance and monitoring (from an infrastructure perspective) can simply be provided to security as a customer of the corporate network, allowing the security director to focus on the business uses for the CCTV system and not the technical maintenance of the system.

Nilsson: There are several benefits that network cameras and IP surveillance solutions offer customers. First and foremost, there is the benefit of lower cost. If it is a new system where a camera has not been previously installed, or a larger existing analog camera system with more than 50 cameras, the system cost is many times lower. An IP surveillance system usually comes out costing even less when you factor in total cost of ownership. Network cameras have additional benefits such as megapixel resolution, the ability to power the cameras over the ethernet, integrate PTZ control, audio, alarm contacts, and more.

Taylor: IP-addressable CCTV offers an inexpensive and convenient method for remote access of the CCTV video. Most IP-addressable systems allow multiple end users to access the same video simultaneously. IP-addressable CCTV also offers methods to allow for flexible permission schemes, so that "power users" can easily be given more rights than standard operators for viewing certain cameras or types of video. A DVMR or IP camera that supports TCP/IP video transmission will likely also provide additional TCP/IP services, such as FTP archiving, e-mail notification on alarm, remote status monitoring and remote control of relays.

Gorovici: A network video recorder (NVR)?not a DVR, which is instead a digital VCR that resides on the network?provides a wide array of bandwidth management and network management tools. It provides multiple solutions in one system?multiplexer, matrix switch, and DVR. It runs on off-the shelf equipment, and it incorporates storage from any of the storage manufacturers. It's also scalable, flexible and upgradeable.

Friborg: IP-addressable CCTV offers many clear benefits, especially if you choose open platform management software from an independent software manufacturer. You will then get a very high degree of flexibility to choose cameras from a variety of manufacturers, to choose the networking equipment and hardware that suits your needs, including wireless equipment, and to integrate CCTV with other systems. Besides this, it's network based, so you can run multiple cameras via the same cable, use Cat 5, fiber, wireless or existing cables, and cover larger areas more efficiently.

Lockhart: The answer depends on at what level the CCTV system is attached to the IP network. The most direct attachment is to a camera that connects directly to the Ethernet and has its own IP address. In this case the camera immediately breaks the old rule of quads and eights and becomes a singular device that can be added to an overall network of cameras in any combination and configuration. When powered by the Ethernet switch, an entirely new concept in deployment is opened by breaking the power tether that long has been the bane of AC cameras. Add the ability to change and customize to requirements and it becomes limited only by the software and storage systems that serve and store the images. In all cases, both digital video and IP transport opens up video surveillance to becoming an ROI resource instead of a direct expense.

ST&D: Are end users and installers aware of the technology and able to implement it effectively?

Taylor: Skill levels in implementing (the systems) are evolving. Although people are aware of the technology, only a fraction of end users and installers have actually installed one because cost, archiving and network bandwidth issues have stalled widespread deployment.

Gorovici: Our experience is that both installers and end users are increasingly aware of the technology, and we are working very hard to make sure they better understand the benefits of the technology and how relevant it is for them today. Just this year, our sales team has given thousands of demonstrations to integrators and end users of every type and level?they're asking us for the demonstrations, and the interest is increasing every day.

Nilsson: End users are beginning to learn about the technology after reading about successful installations in security and vertical magazines. They are seeing that more and more vendors offer the technology. However, there is still a lot of confusion and misperception on the market, created mainly by the fact that DVR technology is still fairly new, and some end users as well as system integrators are just getting up to speed on this technology.

Ian Ehrenberg, vice president and general manager, NICE Systems: We are seeing more and more demand for IP-based solutions, mainly in the transportation sector. This clearly indicates that more end users have bridged the knowledge gap. Moreover, we are seeing end users and consultants starting to seek more value from digital video than simply getting it from the camera to the control room over a network. The ability to detect events, such as perimeter intrusion or suspicious bags or vehicles, dramatically enhances the value of distributed video solutions.

Lockhart: From the end-user perspective, the technology can be broken into two distinct groups of expertise: those who know security needs and those who know networks. As the video security requirements for digital storage and IP transport become more common, the role of the IT/IS discipline will be more dominant. The overall funds for these systems are actually moving under their control now. That said, the IT community has no desire or reason to be security experts and vice versa. A coordination of knowledge of both will be required to understand the new technologies and hardware. The biggest shift in knowledge will occur on the installer side, but then only on the physical installation.

Brahms: Awareness and interest in IP video products seems to be growing. Effective implementation depends on the knowledge level and willingness to learn of the integrator and installer. Satisfaction of the end user is dependent on the product meeting the end user's expectations. If the initial impression is one of technical complexity or if the customer has unrealistic expectations of what the product can do, this can leave a negative impression of the product even if it is performing within its specifications. If the customer's first experience with IP surveillance products is one of frustration, they may be soured on considering similar products in the future.

ST&D: How do you answer an end user who says, "I don't know the first thing about IP/Ethernet. It seems far too complicated to even consider."

Ariel Silverstone, chief information security officer, Temple University: You educate them. Here are the benefits, here are the drawbacks. You do not need to get technical on the issue. Show a basic ROI.

Gorovici: We have seen a fairly high level of sophistication and understanding of what an IP-based system can bring to their operation. The fact is, any mid-size organization has internal expertise to understand this technology, but it's not always in the security department. The key is cooperation between IT and security, and for security to understand they have unique expertise in security management that IT will never have.

Taylor: First of all, anyone who can go online at his or her business already has an IP address. And, with the exception of small businesses, the end users' IT department will be the integrator of the security system to the broadband system and will assign the bandwidth to meet the system's needs. As long as the installer understands the basics about IP/ethernet, the end user does not have to understand it at all. A good system will shield the end user from the underlying technology. End users should look for manufacturer's certification to ensure that their integrator has the sufficient skills to do the job well.

Ehrenberg: This industry is in a state of transition?from analog, to digital, to IP. We recommend to our end-users to look for scalable solutions, which will enable them to shift to IP at their own pace.

Friborg: Within the higher-end market segments on which we focus our sales efforts, we don't get this question so often anymore. Should we get it, it is our experience the customer is willing to broaden his mind when we have explained the main benefits. As the next step the customer's IT department is often involved, and most mysteries can be explained.

ST&D: Some end users say analog cameras always seem to give a better picture on their monitors. Is there some truth in this statement?

Silverstone: Yes. That depends on the line speed, the recording compression ratio, the light conditions and the cameras used. With the right combination, you can easily exceed the quality of analog images.

Telders: Our digital cameras and monitors have significantly better images than the analog system they replaced. It should be noted, however, that not all cameras need to be at the same level of clarity; it depends entirely on the business use of the cameras. One consideration that should be taken is that IP image enhancement has less of a proven track record. There are investigative tools that are well established for analog systems to clean up images and get better results during investigations. Digital systems also have this capability, and the tools are improving all the time. However, with digital technology a pixel is a pixel, and the clean-up process does not have the same degree of established acceptance yet. The solution with a digital system is to design the system for high quality up front so it does not require enhancement.

Brahms: Yes, there is some truth to this statement. When images are digitized and compressed, there is some loss of picture resolution. The more compressed a digital image, the higher the degree of image loss. This is generally true of all the varying image compression methods and particularly evident in network viewing applications, where there are compromises between bandwidth usage, rate of transmission and image file size. Some of the newer advances in compression technology have slightly improved this loss, yielding a better-quality image from a smaller compressed file size.

Nilsson: The belief that analog cameras provide better image quality than network cameras is one of the most common misperceptions in the market. Some early network cameras did not have very good image quality, and that could be the original source of this belief. However, the image sensor (CCD sensor) in a network camera is exactly the same as that in an analog camera. The image quality has been the same for the last two years. In fact, the image quality from a network camera is now actually much better, for two primary reasons. The first reason is the emergence of progressive scan sensors that eliminate interlace problems which occur in analog cameras at 4CIF resolutions. The second reason is megapixel sensors, which are only usable with network cameras.

ST&D: What should physical security end users know about bandwidth to help them talk intelligently to IT professionals?

Nilsson: They should know that it is available and cost efficient and that any system can be accommodated. The largest system we have seen to date is 700 cameras on one site, and the largest total system is 8,000 cameras over several sites. Also, any professional video surveillance system should not attempt to put the surveillance video on the existing office network. The video network should be specifically designed to the video surveillance application.

Telders: Have your installer give you specific information regarding bandwidth utilization. The corporate network may not have the horsepower for extensive streaming video unless you have carefully coordinated with the IT staff. There are many ways to balance the throughput for a system. Have your network and installer staffs work together to design a system that meets both your needs and the network needs.

Silverstone: Just the bare minimum: how wide, switched or not, and acceptable delay.

Brahms: Know the bandwidth requirements of IP hardware and software you are working with. Bandwidth is always at a premium, and this is a question most IT people will ask when discussing addition of new IP products to their networks. Also, most products' performance over a network will vary depending on the available bandwidth. For instance, a browser-based Web camera performs well on the local network, but when accessed from a remote location the image refresh rate is painfully slow.

ST&D: What can an end user do to future-proof their solution when technology moves so fast?

Gorovici: A software-based system with frequent and cost-effective upgrades is the very definition of future-proof, and it certainly beats bringing in the forklift to replace all those DVRs every two or three years.

Taylor: End users should look for a technology vendor who has a long history of innovation and customer service. Much of the innovation that occurs from one year to the next is available in the form of software upgrades, so end users should make sure that their system is capable of accepting software upgrades as they become available.

Nilsson: End users should look for technology where they can re-use as much of the existing system as possible. If for example, 100 analog cameras were installed two years ago, most of the cameras will probably have another couple of years left. Video servers can then be used to digitize those cameras and bring them into an IP surveillance system. End users should also look for products using standard compression, e.g. MJPEG, MPEG2 and MPEG4, which are compatible with several video management systems.

Friborg: The management software is a very important piece, since it ties everything else together; it therefore needs to be flexible enough to suit the needs today and tomorrow. Among other things, make sure you can run the management software on standard computer hardware. If possible, buy management software that allows for the use of both MJPEG cameras and MPEG4 cameras.

Telders: Focus on non-proprietary systems whenever possible. Proprietary systems are rarely as flexible to changes as you would like them to be. It limits your ability to use other components. I like the commodity approach to components as long as you also focus on quality.

Silverstone: Use fiber instead of copper, install multiple storage drives or rely on a SAN-based solution (Yes, share with IT!) and plan on PTZ down the road if not already installed.

Brahms: Stick with products that use standardized compression technologies and access methods. Use hardware products that can be updated via firmware. Consider the use of video Web servers, which allow upgrade of existing CCTV systems for IP transmission and should they become obsolete allow for replacement of a single component rather than the entire system. Most Web servers allow for the loop through connection of analog video signals, allowing continued use of existing quads, multiplexers and recording systems such as DVRs and VCRs. If combined with off-site recording of digital video, this provides a level of redundancy in the system, helping to prevent loss of evidence.

Loading