How to create artificial intelligence to interact with electronic devices

Thread Starter

victoire KALUNGA

Joined Dec 5, 2023
1
Besoin d'aide svp je suis étudiant à l'université de Lubumbashi en République Démocratique du Congo en bac2 , je travaille sur mon projet de dernière année qui parlera - "Développement d'un dispositif de communication neuro-électronique pour les personnes atteintes de paralysie" mon problème se pose au langage utilisé pour ce dernier, car mon ia doit interagir avec des dispositifs électroniques , alors quel langage me proposez-vous ?

Mod:

Need help please I am a student at the University of Lubumbashi in the Democratic Republic of Congo in bac2, I am working on my final year project which will talk about - "Development of a neuro-electronic communication device for people with paralysis "my problem arises with the language used for the latter, because my AI must interact with electronic devices, so what language do you suggest to me?
 

Ya’akov

Joined Jan 27, 2019
8,974
Need help please I am a student at the University of Lubumbashi in the Democratic Republic of Congo in bac2, I am working on my final year project which will talk about - "Development of a neuro-electronic communication device for people with paralysis "my problem arises with the language used for the latter, because my AI must interact with electronic devices, so what language do you suggest to me?
Welcome to AAC.

Choosing a language requires more than the title of your project. Many factors will need to be considered.

  • What is the hardware, and what language(s) does it support?
  • What strategy will you be using to process the data you get from the sensors?
  • What libraries are available to implement the strategy?
  • Are there reasons to use more than one language? (e.g.: best for hardware interface, data processing, user interface, &c.)
  • Which language(s) are you most proficient in?
  • Other things that will be revealed by the design phase of your project.

Aside from the absolutely necessary step of fleshing out what that title means, that is, using something like a user story to describe what the finished project will actually do, here are the steps more or less you will need to take.

Using the requirements in the user story for functionality,

Start by working out what hardware you will be using. You will need sensors and there is a good chance you will choose something with hardware-based processing included. Further, you might find that it makes the most sense to use an FPGA on the front end in which case you will need to use something like HDL for programming that part.

Then figure out how you intend to process the data coming from the sensor (array) and its processing hardware. This strategy might include, for example, Machine Learning. If that’s the case you have to choose a library to use and might choose, say, Python because of the ready availability of libraries with Python bindings and example code.

You might find you have a need for some high performance inner loop code, in which case you might choose C/C++ for that, or even Assembler for the hardware architecture you are using.

Then determine what sort of user interface requirements you will have. Since it is the easiest and most ubiquitous, you might choose to use a browser-based interface—in which case Javascript might be the best option for you. Or, you might even choose Electron to create a non-browser interface, but that would still require Javascript.

So, the bottom line is: don’t engage in premature optimization. You can‘t choose the language independently—you first have to find the parts and techniques you will use, then work out how to connect them together, then finally you will have a much better requirements list to choose language(s) which may not even be necessary as things may simply dictate certain choices on their own.

Unless this is the most trivial project (thinking to turn a light on and off, for example) you have a fairly complex set of discovery, design, and development activities. And, even if it is as simple as that, you still have 60-80% of the same activities. If you want to do this right, front load your project with research and planning, make a genuine design (that is a functional block diagram that specifies subsystems), and only then start trying to build something.

You can and should work up little examples of working hardware and software during the design process to make sure you understand how things work. Development boards, example programs, and the like are good resources for that. While the first parts of the process are relatively abstract, anchoring them to the concrete with little bits of practical circuitry and code will keep you on track.

Good luck, and don‘t forget to start with a robust description of the problem you are trying to solve and what the solution you are building will look like from the user perspective when it is done.
 
Top