Our world is not as it used to be. We have all kinds of stuff happening around us from horrendous crimes to mass killings. To prevent such events, our resources are limited. We are having to rely on a greater number of security personnel and assistance from the military. Despite our best efforts and use of all the possible resources, we are still failing miserably to stop the attacks from happening. So while sitting in the car, waiting to pick up my mom from school, I had this idea. We seem to be better in reducing the damage more than stopping the damage from the occurring the first place.
We seem to be on high alert after an incidence to prevent further damage but we seem to be incapable of stopping all the incidences from occurring. The main problem is not with machines we built, or the technologies we use, or the advancements in technology but with people. If we can stop the people, we can stop the things they do. We all know this already.
With so much data being collected and so many things happening around the world, its almost an impossible task to track, monitor, and analyze everything.
My Solution – Build Swarm of Bots that act collectively. Oh boy, when I have kids, I can so imagine one of them ending up in videos like this while daddy demonstrates to the world how swarm bots work 😆 😆
Anyway, when a friend of mine recently added me on Facebook, instead of just thanking her for the add, I thought I’d begin our conversation by asking her about why they give us tablets when we all know that flu is a viral infection. She is a science student so I was hoping she would know this kind of stuff. She gave a good answer about how antiviral tablets that are usually prescribed aid in producing antibodies that fight the virus and how they prevent other infections. So I thought that was quite interesting how God made antibodies to respond to foreign “stuff” in our blood stream.
After researching a little bit and recollecting things I learned in year 13 biology, I realized we could use this along with some other technologies that exist to respond in a similar way. Our body is under constant attack whether from the outside from the virus, bacteria, and parasites or from the inside for instance with the development of tumor cells in cancer. The body has specific self-defence mechanisms.
For outside attacks, the many antibodies in the body are designed to recognize specific pathogens. They bind to their target and take it to an effector cell which recognizes them through receptors. When several antibodies bind to cell receptors, the pathogen is swallowed up and destroyed. In the event of an internal attack arising from the development of multiplication of tumor cells, the antibodies bind them to the effector cell. The destruction mechanism involves the release of toxic granules that pierce the tumor cells membranes.
In the event of an internal attack arising from the development of multiplication of tumor cells, the antibodies bind them to the effector cell. The destruction mechanism involves the release of toxic granules that pierce the tumor cells membranes. Likewise, my idea was to develop drones that are shaped similar to antibodies in our body. The Y shape allows it to deposit immobilizing content into a human while the top two flaps give it surface area needed to be powered by solar energy.
The drone must be small in design as it needs to fly and most importantly stay without recharge for long periods of time. It needs to sustain bad weather conditions which I am still thinking about. We can make use of charging stations that allow these drones to touch base and keep in connectivity. It will also enable them to land and recharge themselves when they are low on power. Just like the antibodies having different classes – Immunoglobulin A (IgA), Immunoglobulin D (IgD), Immunoglobulin E (IgE), Immunoglobulin G (IgG), and Immunoglobulin M (IgM), we could have various classes of drones each having a different objective but collectively working together in meeting the assigned objective.
Using Swarm Intelligence and Artificial Intelligence techniques, we could create a collectively working swarm of bots that achieve an established set of principals. Swarm Intelligence is the collective behavior of decentralized, self-organized systems, natural or artificial. This expression was introduced by Gerardo Beni and Jing Wang in 1989 in the context of cellular robotic systems. It consists of a population of simple agents or boids interacting locally with one another and with their environment.
The inspiration comes from nature specifically to how bacteria replicate in order to achieve I guess their sole purpose of increasing in number and taking over the host. The agents follow very simple rules, and although there is no centralized control structure dictating how individual agents show behave, local, and to a certain degree random, interactions between such agents lead to “intelligent” global behavior observed in ant colonies, bird flocking, animal herding, bacterial growth, and so on.
In Reynolds paper on boids, He demonstrates an artificial life program which simulates flocking behavior of birds. In that, the rules applied were separation (steer to avoid crowding local flockmates), alignment (steer towards the average heading of local flockmates), cohesion (steer to move toward the average position of local flockmates). One could go on to add a more complex set of rules like obstacle avoidance and goal seeking.
Unfortunately, there isn’t a lot of research on goal seeking but hopefully, the idea is to train through machine learning and video analyze on what would be classified as suspicious behavior and how a swarm of bots acts towards the safety of humans. There are other interesting fields that one could adopt such as Evolutionary Algorithms which has been my all time favorite but more research needs to be done in this area.
Stochastic diffusion search is another agent-based probabilistic global search and optimization technique which will work well for our problem here where the objective function can be decomposed into multiple independent partial-functions. Each agent maintains a hypothesis which is iteratively tested by evaluating a randomly selected partial objective function parameterised by the agent’s current hypothesis.
One other way forward is to use video analysis and machine learning to make a decision on which the swarm of bots responds. For example, a swarm of bots could immobilize a human who is covering his head and has a weapon in his hand heading towards a store at night. The video analysis would pick it up as a burglary in progress and issue a command for the bots to act. A simpler method is to inject an immobilizing content (anesthetic content) into the subject so the police could reach and make appropriate decisions following that.
As I said earlier, there is plenty of research to be done into building an application that could successfully be deployed in real-life. I am still thinking about a lot of these things and the best way to move forward so watch the space folks. I think we are on to something here 🙂
Image Credit: https://www.helpmeoutdoc.com/wp-content/uploads/2015/01/Antibodies.jpg