Agility Communications was established in October 1998 by Dr Larry Coldren and Greg Fish. The company is headquartered in Santa Barbara, California, where an indium phosphide (InP) wafer fabrication plant is also located. A packaging and assembly plant has additionally been established at Allentown, Pennsylvania. The company currently employs 200 staff, many of whom are engaged in manufacturing.

Commenting on the staff numbers, Mr. Martin noted that the high number of manufacturing staff is largely due to the necessity of establishing an InP fabrication plant because of the current lack of companies to whom such processes can be outsourced. He added, "Perhaps in ten years time, a company such as Agility Communications would be able to outsource laser production processes and so set up with substantially less funding. Barriers to entry in this market sector currently are high.

The company has raised over $200 million in funding over four rounds to date. The most recent round closed this February with an addition of $27 million to the round opened in August 2001, with $83 million. Strategic investors include Siemens, through Mustang Investors, Ciena and Tellabs. Arlon Martin stated that Agility has substantial cash reserves at present, which are projected to be sufficient to sustain the company for two to three years.

Outlining the problem, Arlon Martin said, "Lasers cost between $1,000 and $2,000, whilst the board that these components go into will cost anywhere from $15,000 to $30,000. Because each DWDM laser is unique, an 80-channel system requires 80 unique circuit boards. The result is that system manufacturers had to stock a wide variety of boards in large numbers in order that they have the appropriate type for each customer. Further complicating this issue is the fact that carriers usually deploy equipment, and capacity in terms of channels, in phases as demand in the network increases. Agility Communications' widely-tunable laser product solves this inventory problem."

The product will also play a role in the execution of dynamic network provisioning, the implementation of which the industry is currently investigating. Dynamic provisioning will allow carriers to change the number of channels 'on the fly' in order to balance the network to accommodate fluctuating traffic patterns. This technique will effectively increase the capacity of a network, without the addition of network components, through the reuse of channels. Network build and operating costs are also reduced by "50% or more", according to Mr. Martin.

Positioning

Discussing the relative merits of widely-tunable lasers as compared with narrowly-tunable devices, Mr. Martin commented that the latter product would largely address inventory issues, although it lacks the versatility and functionality of the former. He said, "Several companies have narrowly-tunable devices with an 8 nm range, capable of switching over perhaps twenty channels. Our device has a range of 40 nm, sufficient to cover the entire C-band. With this capability comes increased functionality - for example fast switching and 'ring-hopping' of traffic. Our product is effectively future-proof."

Agility Communications is solely engaged in production of widely-tunable laser devices. To date, the company has seen no value in producing narrowly-tunable devices, as the cost to the customer for such devices is the same for less functionality. However, it is possible that narrowly-tunable devices would be offered if there was found to be customer demand for them.

Putting into context the development of tunable lasers, it was noted that the product has been promoted for several years. Arlon Martin was asked why has it taken so long for the devices to reach acceptance with the systems manufacturers. He commented that laser architecture transitions have always taken a number of years, and the same is true for fibre amplifiers - both 1400 nm and 980 nm pumps.

According to Mr. Martin, "There was a lot of discussion in the industry concerning reliability of 980 nm pumps before Lasertron gathered enough data to persuade some key players that reliability was not an issue. This then led to the industry as a whole adopting 980 nm pumps, to the point today where these components are installed throughout the entire telecom space, even submarine networks."

Arlon Martin believes that the same scenario will hold true for tunable lasers. Development of these devices has only come to the forefront for systems manufacturers over the past two years. Up until that time the only company active in this area was ADC, a competitor of Agility Communications. Smaller companies, for example Atoga Networks, are incorporating tunable lasers into their products, but the major players have not publicly announced their plans.

Mr. Martin commented that smaller systems vendors have the advantage of starting with a clean sheet - they can assess the available technologies and select those that they see as the best fit for their particular strategy; large vendors will have a lot of history and 'baggage' to accommodate.

As a proviso, Mr. Martin added that it is never exactly clear as to what areas the large systems manufacturers are working in, but it is certain that at least some of the major players are in fact developing products incorporating tunable lasers. In support of this statement, Mr. Martin said that although he was not at liberty to reveal the names of companies engaged in projects with Agility, it should be noted that Ciena, Tellabs and Siemens all made substantial investments in Agility Communications during the last funding round.

Ultimately, equipment vendors must continue to develop flexible systems that reduce capital expenditure and operating costs for their customers, the carriers. As mentioned previously, tunable lasers enable remote provisioning of the network from a computer terminal. Arlon Martin commented that in today's networks, without tunable lasers, this provisioning requires that a technician visit the relevant node in order to install a new board - a procedure that typically costs in the region of $100,000.

Distribution

Agility Communications has, to date, employed a direct approach to sales. Mr. Martin said that the sales process for its type of product is highly technical - expertise that is often lacking with distributors. Furthermore, tier one customers will generally expect to deal with a supplier such as Agility Communications directly.

Partners or distributors will be used where appropriate, for example when entering the Japanese and Chinese markets. The company recently announce Nissho Electronics Japan as it distributor for the Japanese market.

The company has established a sales team in Europe, headed by Managing Director James Regan, previously with Nortel Network's Optical division in the UK.

Describing this design, Arlon Martin explained that, generally, lasers function by use of a cleaved facet, whereby the mirrors allowing the chip to lase are formed when the chip wafer is cut. Agility Communications' laser is unique as the two mirrors, composed of indium phosphide, are integral to the chip - resulting in a longer than normal chip. The laser is tuned by varying the electrical current applied across the chip, which changes the length of the cavity within the chip. This technique gives a very wide tuning range.

Intellectual property

Agility Communications has licensed a number of patented technologies developed at UCSB. The company also has further patents pending for technology developed since Agility Communications was founded.

Agility Communications counts a number of customers who are either beginning to sell systems into the carrier market, or designing systems in their laboratories, incorporating its devices. The major players in the optical market, such as JDS Uniphase, Agere, and many of the Japanese companies, do not yet have access to a widely-tunable chip architecture. At the present time, the transmission laser market is controlled by the likes of Agere, Nortel and JDSU - with fixed wavelength lasers.

Two companies focusing on widely-tunable laser devices are Bandwidth9 and iolon. Bandwidth9 uses VCSEL technology - in contrast to the in-plane technology of Agility - in the production of low cost lasers for the access market. Bandwidth9's devices have the disadvantage that they are of low power - at 0.1 mW, around 100 times lower power than a traditional fixed wavelength laser - so restricting their use without the addition of expensive fibre amplifiers and wiping out any cost advantage.

iolon produces devices based on a Fabry-Perot laser, which utilises an external cavity approach. With this device, the laser is directed onto a grating, then to a mirror, which reflects the light back into the laser before it is transmitted along the fibre. This design carries the disadvantage that moving parts are involved - the grating and the mirror - increasing production difficulties and therefore costs. Commenting on iolon's approach, Mr. Martin said that such devices would be expected to have a place in the market, once the difficulties of manufacture have been overcome and any reliability issues have been addressed.

Agility Communications' widely-tuneable laser device has the power of a conventional fixed wavelength DFB device - approximately 10 mW. The device incorporates no moving parts and employs electrical current to alter the wavelength. An additional benefit arising from this technique is that switching is achieved extremely quickly - within tens of nanoseconds. Commenting on this, Mr. Martin noted that mechanical switching is comparatively slow - often barely meeting SONET specified restoration times.

Agility Communications' devices are based on a single chip and packaged into a fourteen or twenty four pin butterfly component - identical to that for a fixed wavelength DFB or DBR laser from any other manufacturer. Hence, Agility's device may be designed into systems in place of fixed wavelength lasers.

With the exception of Fujitsu, major Japanese players are generally absent from the tunable laser market. Mr. Martin noted that these companies had missed the boat during the transition to DWDM systems in the mid-1990s. One factor hampering the Japanese manufacturers was the installation in Japan of 'the wrong sort of fibre' - dispersion shifted fibre - which effectively excluded the use of DWDM systems. Whilst their North American counterparts were developing 2.5 Gbit/s and 10 Gbit/s multi-channel optical systems, in Japan companies were producing 10 Gbit/s TDM equipment. As a result, the Japanese manufacturers lost ground to their North American competitors in terms of laser, filter and multiplexing technologies.

Arlon Martin added, "In 1993 to 1994 there was a big debate in the optical industry over whether network capacity in excess of 10 Gbit/s over a single fibre was needed. In North America a multi-channel approach was adopted - initially systems offered eight 2.5 Gbit/s channels, which could be upgraded to sixteen channels, and so on. Ciena and Pirelli were the first companies to launch DWDM systems onto the market. Lucent quickly followed with a 2.5 Gbit/s 16 channel solution. Subsequently, Nortel launched a 10 Gbit/s multi-channel system. At this time, in Japan, 10 Gbit/s TDM systems were being developed - there was not a market for DWDM equipment."

The Japanese players are now getting back into the game, with Fujitsu in particular seeing significant successes in the North American market. Mr. Martin noted that these companies - notably Fujitsu, Hitachi and NEC - are developing a strong presence in the optical components market and remain a force to be reckoned with.

Pricing

Presently, Agility believes that it can supply its widely-tunable lasers at a premium of approximately 20%, compared to the price for a fixed-wavelength device. This pricing comparison is based on current street prices for lasers ranging from around $500 to $2,500. The company anticipates reaching price parity with fixed-wavelength lasers by 2004.

Arlon Martin commented that, at the circuit pack level, tunable lasers would add only a few percent to the overall cost. On the other side of the equation, Mr. Martin reiterated the point that system vendors would realise substantial reductions in inventory costs through the use of widely-tunable lasers.

Arlon Martin said that currently fifteen vendors worldwide have been targeted; this figure is not expected to climb to more than thirty over the next two years. Explaining this strategy, Mr. Martin stated that this is working on the theory that many smaller systems vendors in the market today will fail over the next few years.

Agility Communications does not view the in-house research and development capabilities of the major systems vendors as posing a threat to the company's future. Speaking about his experience when working at Lucent, Arlon Martin said that these major companies would always buy-in certain technologies and capabilities. Partnerships and strategic alliances are a likely route to selling into the large systems vendors for Agility Communications, through, for example, participation in multi-sourcing agreements.

Customers

Agility Communications has publicly announced two customers for its product. Atoga Systems operates a ring architecture, into which tunable lasers have been installed at the nodes, alongside fixed receivers. Agility's widely-tunable lasers allow communication with any receiver point on the network, at any time a receiver is open. The result is an ability to communicate between any two points on the network, regardless of what traffic is being carried by the network at the time. Agility expects the market for this type of application to take off over the next few years. Agility is also working with Network Photonics.

A further application for the company's devices is at SONET metro ring intersection points, to enable communication between nodes on separate SONET rings. This process requires that the wavelength of the signal be altered enabling traffic to be 'hopped' from one ring to another. Achieving this using fixed wavelength lasers would require the installation of a rack of lasers covering all possible wavelengths.

Mr. Martin did state that the company has entered into discussions with vendors accounting for 80% of the DWDM equipment market and that further announcements of customer relationships would be forthcoming.

He added, "All of these companies are investigating the use of tunable lasers in their systems at some level. Only one sector of the market is likely to be closed to us - the access network. In the metro, long-haul, indeed any network where carriers have a requirement for switching multiple channels, tunable lasers will be embraced to some degree."

Market size

Commenting on Agility's assessment of the value of the total market for transmission lasers, both fixed and tunable, Arlon Martin cited a figure of $1.2 billion, worldwide, for 2002. Mr. Martin noted that this figure represents approximately 30% of the market at its peak in 2000. 2002 is anticipated as being the trough for the transmission laser market, which is expected to climb to around $2 billion by 2005.

Mr. Martin said that the widely-tunable laser portion of this total market cannot currently be calculated due to the fact that systems incorporating the devices have not progressed beyond the development stage. He added, "Atoga Systems themselves, who are utilising our product, do not yet realise revenue. Projecting forward, I would expect widely-tunable devices to account for around 50% of the total transmission laser market by 2005."

Arlon Martin proffered the view that there are no effective measures for market development and growth within the telecommunications industry. Market size calculations and estimates are, ultimately, pure guesswork. This is in contrast to, for example, the semiconductor industry, for which there are accurate measures of unit shipments, enabling reliable estimates of growth and giving a realistic picture of the market. Equivalent base data for the telecommunications industry simply does not exist.

Putting this comment into context, Mr. Martin explained that Agility Communications had taken as its leading indicator fibre sales from Corning - the supposition being that, as the market slowed, carriers would stop buying and installing fibre to concentrate on connecting network equipment to fibre already in place. It subsequently emerged that this measure was baseless, due largely to the fact that systems vendors, such as Lucent and Nortel, were using vendor-financing packages to push surplus product into the market. This situation was compounded by the exaggerated forecasts for network equipment requirements passed to the systems vendors by the new carriers, primarily the CLECs. These overblown figures then fed back from the systems vendors to the components suppliers.

On a positive note, Arlon Martin added that, even discounting the erroneous numbers, the demand for telecommunications equipment continues to grow. He stated, "There is nothing fundamentally wrong with the business, even if 2002 promises to be a challenging year."

Referring to the 980 nm pump market, mentioned earlier, Arlon Martin noted that Lasertron, and later SDL, were able to gain access to the major equipment manufacturers with their 980 nm pumps because they had a product superior to that developed by the big manufacturers in-house. "The same will hold true for the widely-tunable laser market. I am confident that Agility Communications can succeed in the marketplace."