M2M helps zoologists protect black rhinos and penguins
May 20, 2015
From black rhinos in Kenya to penguins in Antarctica, an M2M camera system is helping zoologists preserve wildlife and continue the fight against poachers. Jonathan Pallant, principal engineer at Cambridge Consultants, explained how this all came about at last week’s Device Developers’ Conference in Cambridge, UK.
It started at a Camjam, a coming together in Cambridge of Raspberry Pi enthusiast during which a representative from the Zoological Society of London (ZSL) explained a problem that its staff were facing. They were running a programme called Instant Wild for tracking endangered species. Cameras placed in the wild would collect images remotely and upload them automatically.
The scientists were using off-the-shelf rail cameras and GSM technology for the communications. Each image was sent as a picture message.
“The problem is that animals such a s black rhinos do not live in areas with good cellular coverage,” explained Pallant at the conference. “Also, ZSL are very good at animal preservation but do not have a lot of funds, so spend them wisely.”
Cambridge Consultants reckoned they could solve this problem using the Iridium satellite network. There are 66 satellites in low Earth orbit. These are all working fine even thoughthey are due to be replaced soon; the next generation satellites are already being built.
When Iridium started, it used a Motorola handset, but the company soon moved to one designed by Cambridge Consultants, hence the relationship between the two companies.
For ZSL, this meant global coverage. Cambridge Consultants designed the three main elements of the new system – the central image server, a node linking the cameras and the wireless cameras themselves. The cameras transmit using RF over the short distance to the central node.
Within the camera box are two cameras, one using infra-red for night vision and a standard one for daylight. They use cheap off-the-shelf batteries that can be bought anywhere as these are being installed and maintained in remote places. One set of batteries will last about two months in normal operation.
The electronics inside the box include an Atmel Xmega processor, Conexant digital signal processor, a Texas Instruments CC1200 as the radio and Omnivision OV7740 sensor. These transmit images to the central node, which is based around a Raspberry Pi single-board computer. Typically, five to ten cameras are linked to one central node, though it can take up to twenty. The node uploads the images via the Iridium satellite network. The Raspberry Pi is running Linix.
“We did field trials in an actual field at Whipsnade Zoo,” said Pallant.
The fact that the whole system can handle temperatures down to -40ËšC led to one of its early deployments, on the Yalour Islands in Antarctica monitoring the movement of penguin colonies. Here, the cameras only take four images a day and so survived for nine to ten months on a remote island. And it meant that the penguinologist knew within four hours when the penguins had returned to the island.
A more long-term deployment is in Kenya where the technology is being used to track poachers hunting black rhinoceros.
“Even I don’t know exactly where these cameras are, said Pallant, “as their location is kept secret. They are camouflaged.”
Altogether five systems have been installed, including in Sri Lanka and the Himalayas.