In the winter of 2003, a solar-powered Web camera was deployed in northern New Hampshire, and in the process, the highest wireless network (WiFi - 802.11b) in the northeast United States came into operation across 5 miles of backcountry. Given the interest in this project, the information below is being provided to answer questions of how the project came to be, the equipment selected, and technical issues addressed.
Atop the Northeast United States highest peak is the Mount Washington Observatory, a non-profit weather research facility. The Observatory has a frame-relay link to the Internet, which is used to disseminate images and video from the Web cameras pointed out its windows at neighboring peaks and regions. One of Mount Washington's most striking features however cannot be seen from any of the Observatory's windows. Tuckerman Ravine, on Mount Washington's southeast side, is renowned for spring skiing, and the 2-3 hour backcountry hike required before carving turns. This project was started to provide near real-time images of Tuckerman Ravine and its snow conditions.
The first step in this project was identifying the best vantage point of Tuckerman Ravine. East of Mount Washington, across Pinkham Notch, is the Wildcat ski area. At the summit of the ski area, there are wonderful views of Tuckerman Ravine, Huntington Ravine, Mount Washington, and Mount Adams, New England's second highest peak. With a site identified, the next step was to determine logistics. In order to pursue the site, permission had to be provided by both Wildcat and the US Forest Service, which oversees the land the ski area is on. Once a tentative go ahead was provided by both organizations, equipment selection could begin and technical logistics could be addressed.
In order to take advantage of the expansive views atop Wildcat, the Canon VC-C4 Pan/Tilt/Zoom (PTZ) camera was chosen. The Canon's 16x zoom capability allows amazing close up shots of the ravines and other features on Mount Washington, while at the same time taking some beautiful shots of the mountain as a whole. The panning and tilting capability has also come in handy in providing views of Mount Adams as well as the ski trails and lift atop the ski area.
When the camera was chosen, there were no reasonably priced PTZ cameras with good optics and built-in Ethernet capability. In order to connect the camera to the network, the AXIS 2401 Video Server was selected. The 2401 video server has an Ethernet port for networking, an RCA jack for video input from a camera, and an RS-232 port to serially control the camera. On the software side, the 2401 video server comes with a built-in Web server that includes CGI interfaces to control as well as capture and stream video from the camera over the Ethernet connection.
Given the lack of communication lines atop Wildcat, a wireless solution was the only viable alternative. As images and potentially video were to be transmitted, a relatively fast connection was required. Additionally, distance, weather, and cost were also a factor.
Atop Wildcat, there is no visibility to the ski area facilities at base level. The only permanent structures visible with direct line of sight are the towers and buildings on the summit of Mount Washington. As the Observatory building has a frame relay connection to the Internet, this was a logical choice for the other end of a wireless network, although it would require a solution capable of spanning 8 kilometers (5 miles).
Weather is a critical factor on and near Mount Washington, home of the world's worst weather. The mountain's climate can rival that of Antarctica, and is where the highest surface wind speed on Earth has ever been recorded at 231 miles per hour. Rime ice constantly covers the exposed equipment in winter, and along with the wind has been the cause of many broken antennas. Even on Wildcat itself, approximately 2,000 feet lower in altitude, there is considerable snow and ice. A wireless solution in these conditions needs to be robust and reliable.
With a limited set of funds for deploying this capability, and with the knowledge that a solution to power the equipment would take up most of the budget, WiFi (802.11b) technology was selected. A diverse set of 802.11b products had been available for a couple of years, at the project's outset, for both the consumer and enterprise market, providing a lower cost alternative to other wireless technologies. The bandwidth provided by 802.11b was certainly sufficient for this application, however the signal degradation due to weather and distance were still a concern.
Given the robustness of the Orinoco 802.11b line, the OR-500 outdoor routers were selected along with 14-dbi directional antennas; this set up is also available bundled as the Orinoco Point-to-Point Backbone Kit. Although the original intent was to use an omni-directional antenna to support possible future deployments of additional remote cameras, calculations of the system operating margin proved to be insufficient to support this application without the use of an amplifier. The use of an amplifier was not recommended because of potential interference with other local wireless uses including search & rescue and government relays. Also, the profile of omni antennas was not advisable because of weather both atop Wildcat and Washington. Similarly, the use of parabolic grid antennas was not considered because of potential ice build-up. As a result, yagi antennas were selected.
Initial outdoor testing of the Orinoco routers and temporarily tripod-mounted yagi antennas on site, yielded a fairly strong signal with bandwidth of 2Mbps. However, concern with the exposure of the antenna on the summit of Mount Washington led us to explore alternative mounting options, and to identify a corner of the Observatory building that had window visibility to Wildcat, readily available power, and where an Ethernet connection to the Observatory network could be run. Testing with the Washington antenna behind (hurricane-proof) glass showed a degradation of the signal, but still maintained an acceptable bandwidth of 1Mbps with minimal packet loss.
On the Wildcat side however, the only structure with line of sight that could house the antenna behind glass, had its windows covered with packed snow most of winter. Above this structure is the old Wildcat gondola bay, which although open to the environment would at least provide some protection and a higher ground clearance for the antenna. This was also the most ideal location to house the camera.
In order to house and protect the camera, video server, and wireless router on Wildcat, a protective box was built and mounted inside the old gondola bay. A cut out on the box covered with a half-dome provided the camera with the desired view. The antenna was mounted directly above the box, ensuring minimal signal loss due to the short antenna cable run.
With no power lines run to the top of Wildcat, a generator is used to power the ski area equipment when the mountain is open and the lift operating. The ongoing costs of running the generator when the lift is not operating and refueling when the mountain is closed in spring and fall, led us to explore a self-sufficient power solution.
Wind and solar power were the logical choices. A hybrid solution would have been optimal as Wildcat may only receive 4-5 hours of direct sunlight daily during the winter. However, with the ski area being on US Forest Service land, little if any visible impact could be made to the ridgeline. This immediately eliminated wind power as an option, leaving only a solar power solution. Although originally intending to place the panels on top of the old gondola building, the panels were instead installed on the south-facing backside of the building in order to have no impact on the ridgeline whatsoever.
The expense of a solar solution limited how long the equipment could be powered. As there was no point in running the camera at night, a photocell is used to keep the equipment load turned off when there is no daylight. In order to further conserve power, a timer is used to cycle the equipment on for three minutes and off for fifteen. The three-minute window is sufficient to control the camera and retrieve numerous images, satisfying the requirements of this project. Enough power is conserved with the use of the timer to allow the future addition of weather instrumentation and a sensor interface to the system.
At the time of this write-up, the system has been under operational testing for two months during the winter of 2003, one of the coldest and snowiest in recent memory. The wireless system has performed flawlessly, even in the worse of weather. Occasionally the camera dome will be covered in snow or ice, however when this happens the outside visibility is limited anyway. On the whole, the project so far has been a tremendous success, having come in under budget and overcome some difficult technical and logistical hurdles.
Visit www.mountwashington.org/cam/ravines/ for a view.
Read Zakon Group Press Release
Read Mount Washington Observatory Press Release
We would like to thank the Mount Washington Observatory and Ossipee Mountain Electronics for having contributed information to this write up.
Zakon Group LLC provided technical guidance in the design, integration, and testing of the system, as well as software development assistance.
Additional Project Photos
Partial Equipment List