South Africa has by far the largest population of rhinos in the world and is an important country for rhino conservation. However, more than 1,000 rhinos were killed last year as a result of poaching, with gangs now moving to neighbouring countries. Intelligent CIO hears from one woman who is trying to combat poaching thanks to Artificial Intelligence.
Colleen Glaeser is the Global Marketing Director for AxxonSoft, one of the largest tech security companies in the world.
She is also the owner of South African luxury safari lodge, Karkloof Safari Villas, and has seen first hand the carnage rhino and elephant poaching can have. Therefore, she has innovatively combined both AxxonSoft and Karkloof to combat it.
“The technology utilises the latest surveillance and security solutions, which incorporates Deep Learning, which is a method under Artificial Intelligence (AI). This technology has been adapted to tell the difference between humans and animals.
“Prior to the incorporation of Deep Learning in anti-poaching surveillance, software often failed control rooms and response units in that false alarms were on many occasions, set off by animals, insects and weather. Control rooms were not able to tell the difference between an actual threat and a false alarm, which often resulted in exhausting resources as teams were dispatched for animals who had touched the fence while grazing in their natural habitat.
“AxxonSoft’s Deep Learning solution now alerts the operators in the control room to an immediate poaching threat as poachers try and breech the fence perimeter to enter the reserve or park.”
Taking around six months to rollout and implement, the technology has been widely received.
“There are rangers on the ground, but they can’t be everywhere especially when you’re dealing with 2,000 acres,” said Glaeser.
“And then you have another complication where the animals might bump the fence and set off an alarm that the electric fence is being disturbed.
“However now, if an alarm goes off, it will only pick up a human and not an animal. We’ve had extremely good success with some arrests being made. The rangers feel more equipped because they know where to go and it won’t be a false alarm. You want to see what type of ammunition people are carrying and what type of arms. And that can only done through surveillance.
“The technology has really been grasped by farm and park owners and we’re currently in negotiations with some of the major parks to deploy this type of surveillance. In the past, they had nowhere to go and didn’t know what they were dealing with. Poachers come well-armed and with a lot of technology behind them, like drones they can access, whereas the park owners don’t have as much to combat it. So, this technology really helps.
“It provides a proactive solution to surveillance whereas previous systems were somewhat archaic and reactive in their response to real threats. Our Deep Learning technology has been extremely successful thus far in telling the difference between animals and humans.
“AxxonSoft has developed a process for adapting neural network filters to the needs of a specific site. The neural network learns to perform customer-specific tasks from video material obtained at the site, which guarantees high-quality results. To resolve issues of high resource consumption related to AI, AxxonSoft uses a blended approach with a neural network filter applied to an object tracker — the filter can identify a specific type of moving objects or abandoned items.”
The technology is available in the Intellect PSIM and the Axxon Next VMS. The classic object tracker detects moving objects or abandoned items in the FoV, before it passes the section of the frame with the object to the neural network for analysis. Any objects that don’t match the defined type are ignored by video analytics in that they don’t trigger an alarm.
“This technology is applied to video that captures high movement from interference (i.e. beyond the object of interest), such as trees blowing in the wind, light reflected off water, traffic and machinery,” said Glaeser.
“Usage examples are alarms that are activated when people appear in hazardous areas of a facility, or applications that reduce the number of false alarms for abandoned objects in public spaces.”
AxxonSoft has also implemented a process for tailoring deep learning video analytics to the requirements of a specific site. To guarantee high quality results, the neural network learns to perform customer-specific tasks from on-site recorded videos.
Glaeser felt that she needed to use her knowledge and expertise within the industry to help curb rhino poaching.
“I see myself as a normal businesswoman but have climbed the ladder pretty high in terms of this industry and have become a sort of thought leader where people approach me in South Africa all the time in how they can use what we’ve done and how they can apply the technology,” she said.
“I know there are a lot of foundations that support the prevention of rhino poaching but my main goal is to have safe parks and nature reserves, so that when people are travelling, they ask if this is a safe park, do you have the type of analytics that are protecting the wildlife and animals and have the owners taken all the necessary precautions to prevent poaching?
“It’s an atrocity when you’re on the ground dealing with this. We’ve now perfected this and it’s working. We just need to raise awareness that the technology is out there, through AI and Deep Learning and we’ve adapted the technology and made it into a smart park.”
Glaeser says her team has worked with various countries within the Southern Africa Development Community (SADC) and the company is looking to expand the technology into other parts of the continent.
“It takes a team to make this work,” said Glaeser.
“From the development team at AxxonSoft Global, to the rangers and control room on the ground, we all come together to make these projects a success. Furthermore, we have over 300 staff members within AxxonSoft and 5,800 partners across the world.
“I do believe hunting has a place in the ecosystem if it’s done for culling but to just to come and shoot an animal because their horns are very big and they’re going to hang it up on their wall is
not for me. I’m not against culling if it’s necessary but I am against shooting just for pleasure. Some of the countries in Africa are third world countries, so it’s about reaching out to them.”