We have the simple Artificial Intelligence to secure our rail tracks

By Hamid Al-Hassan Hamid

I wrote twice about possible attacks on our rail tracks; it is just a matter of time. This, in my opinion, is just a test run; expect more to come if we continue to neglect simple and sincere advice due to ineptitude and corruption. The rail tracks are not left alone, on their own, anywhere in the world. They are protected, monitored and secured. It is done through determination and sincerity of purpose. How many souls would have been lost had the rail skidded and crashed! How disastrous!

Again, with all our tech universities, we cannot build local drones to fly 24/7 and monitor at least our rail tracks. The only thing our professors are good at is attacking another person who became a professor that they do not like.

The technology we need to curb these security challenges is too expensive to buy; we do not have the money. But it is cheaper to develop, and we can do that locally.

I once reached out to MTN, asking how much it would cost me to connect drones that will fly across the country, especially our forest, for intelligence gathering. I will build the server, and they will provide the network, without the Internet data, I don’t need the internet. They gave me two options:

1. Pay 150,000 naira monthly to connect as many drones as possible nationwide.

2. Make them partners in the project, and I will not have to pay a dime.

They needed confirmation and approval from appropriate security bodies. It has been about a year or so now. Getting the interest of the appropriate security bodies alone is more complex than quantum physics.

In Africa, the only thing we love is physical cash, but I don’t blame us. I just pray that God cures our sickness soon.

We need to establish private tech defence companies that are private entities and not owned by the government.

Artificial Intelligence has more practical use cases in Africa. In addition, it will be easier to implement because the biggest fear against Artificial Intelligence is that it will compete with humans in jobs and take away those jobs.

Africans don’t want jobs; they just want to have something to eat throughout the week. Forget about the rampant cry of unemployment. As soon as you employ, you will begin to see. Artificial Intelligence will have no resistance in Africa, especially in security.

What shall we do?

I have been getting messages and comments from brothers trying to help with the Private Defence Tech Company Startup. Some proposed sending proposals to either Minister of Communications and Digital Economy Professor Ali Ibrahim Pantami or Vice-President Yemi Osinbajo. Some also proposed promoting the idea in media houses until it reaches the ears of those in power.

First of all, I have access to Professor Pantami through childhood friends who can meet him whenever they want to. I also know people that can reach the VP. But I disdain the idea of sending proposals.

This is what I am doing at the moment:

I have a team of four individuals with backgrounds in the military and tech. We are making plans to partner with anybody (with genuine sincerity) interested in starting something simple that can be pushed into the market for testing and continue building from there.

At this point, what we want is to partner with the research department of any Nigerian university, military institutions like the Airforce Institute of Technology (AFIT), or the Nigerian Defence Academy (NDA). We want to start by building an AI that will be able to:

1. Identify faces at entrances through cameras.

2. Log check-in and checkout time of each face.

3. Determine if anyone checked in and did not check out, and report such cases to analyse the data to know why such checkout did not occur. The checkout may be missed because the camera did not capture the face or another exit was used, in which case we would like to know if the use of such alternate exits is valid and improve the AI to be more accurate with regards to missing faces.

4. Print out daily, weekly, monthly and yearly statistics on such check-ins and checkouts per individual and the whole entrants.

5. Try predicting possible movements of each individual based on the data collected as they grow.

6. Send silent alerts to mobile phones of respective security personnel on duty if a breach in the entry is detected, for example, an individual using an entrance or exit that is not within their jurisdiction.

We can develop the AI, create the server, and assist with the statistics as part of our responsibility.

We can start by using cheap Android phones as cameras at respective entrances and exits by connecting them to the server via wifi; this cuts down costs by far at the initial stage.

We want to grow the system gradually by later introducing drones to fly outside and see if they can recognise personnel that have been logged in the building at various entrances, identify the cars they use, log their car plate numbers, identify what canteen they like taking coffee within the vicinity and so on. Then gradually scale to state and federal levels.

It is very simple. But can corruption and corrupt individuals allow this?

Hamid is a social commentator, an expert in AI and writes from Sudan.