The timely operation of street lights, porch lights and other forms of outdoor illumination is required for the twin concerns of safety and environment protection; lights operated too late are a safety hazard, while lights operated too early are a waste of power. The switching on and off of external lightings can be set with the help of the device called Twilight Switch.
Traditionally switching has been controlled either via a human operator; automatic control is accomplished using a light sensor like LDR or photosensitive cells. Unfortunately the use of sensors is accompanied by wastage of power and unreliability because sensors invariably deteriorate over time due to exposure to rain, dust, snow and hail, degradation of materials due to sunlight, and other elements like littering and picking by birds.
Three students at NSIT, Akash Verma, Eklavya Gupta and Nipun Garg, pursuing the Bachelor of Engineering degree specialising in Instrumentation and Control Engineering, have developed a twilight switch under the able guidance of Dr. Dhananjay V. Gadre.
The Twilight Switch is a solution addressing the above concerns; it automatically calculates everyday the start and end of civil twilight for that day using accurate mathematical formulae and the geographic co-ordinates and current date as input data. Since the lightings are turned on and off when it is required, electricity is saved. No attendant is needed to perform this function. By being fully automated, this device adds economic viability by eliminating the requirement of an operator for switching on and off lights, only requiring a one-time configuration at the time of installation. By being devoid of sensors, it is possible to enclose the switch in a box away from environmental hazards. This model is an improvement over previously existing ones because of its ease of operation, ruggedness, performance in terms of accuracy of calculated times and because it is a low-cost model.
In terms of calculation of operation times, the Twilight switch gives results accurate to 2 minutes for all locations on the earth below 80° N; a maximum error of up to 10 minutes for the extreme case of twilight approaching 24 hours or nil duration. This error is observed on 3-4 days in the year for high-latitude regions. Indeed these "errors" should be called "deviations" from the standard used for testing; actual twilight times could easily deviate by bigger margins from the standard owing to a multiple of reasons like local topography, weather, temperature, air quality etc.
A prototype has been kept in operation since the 10th of August, 2010 at CEDT Lab at Netaji Subhas Institute of Technology and the operation of lamps at dawn and dusk has been as expected. The problem of drift in frequency has been brought down to -1 second per day, as mentioned before, and the behavior is logged periodically. The system clock on the kit has shown stable behavior and a negligible error, and the last problem - that of drift in frequency of crystal - may be considered resolved.
This project has some more scope left for future research in this field. The effect of temperature on clock accuracy can cause considerable error to accumulate over time in conditions of extreme climate, requiring more frequent recalibration. The use of temperature controlled crystal oscillators for minimizing this need of recalibration is being investigated for use in such harsh climates.