We only have one planet, which is why each of us has the duty to do your part to safeguard it and keep it healthy for as long as possible and this also includes the treat technological waste correctlylike the old smartphones, which crowd our homes. In carrying out this last task, we could soon have an extremely technological help: that of a robot with artificial intelligence in charge of disassemble the smartphone in order to recover all recyclable parts.
A new project: robots with AI to recycle smartphones
In the meantime, a lot of time has passed, the giant of the bitten apple continued to talk about its commitment to the environment and continued to develop its own robot (the latest generation one is called Daisy) to recycle iPhones, however, never made public a crucial data: the number of devices these robots actually recycled to recover reusable components.
Very soon, however, this need for greater clarity regarding the potential impact of robots equipped with artificial intelligence on our ability to recycle technological waste, the so-called e-wastecould find satisfaction thanks to a new research project. The purpose of this study is precisely to realize tools with AI that allow a robotic recycler to recover useful parts from many different smartphone models.
By the time this ambitious technology is commercialized – and here, sadly, we are still forced to move in the field of hypothesis – researchers would have a mine of information at their disposal to improve our ability to recycle smartphones and other small electronic devices.
New challenges and a lot of work to do
At present, e-waste recyclers are mostly dealing with large appliances, such as CRT TVs, however they are more and more smaller devices like smartphones and tablets in front of them.
This new reality presents us with new challenges for two main reasons: this e-waste is often difficult to disassemble and takes much longer. For this reason, most of the time recyclers simply remove the batteries and throw everything else away, thus wasting potentially reusable components such as motherboards but also valuable materials (not to mention the energy needed to produce those components. thrown away and that to produce new ones).
For years now, researchers have been studying the possibility of exploiting artificial intelligence to reverse this trend and perhaps we have come to a turning point: Last December, the US Department of Energy put on the table a $ 445,000 grant for researchers at Idaho National Laboratory, University of Buffalo, Iowa State University, and e-waste recycler Sunnking to develop software that allows robots to automatically identify different types of smartphones to recycle, remove their batteries and recover valuable components. Following a two-year research project, the team hopes to begin field tests with the first versions of this technology at one of Sunnking’s facilities. The next step would be to raise more funds for bring robotic smartphone recyclers to the market.
The project: details and opinions
Amanda LaGrangeCEO of the e-waste recycler TechDumpspeaks of a work of fundamental importance to improve the sustainability of consumer electronics, but also shows some skepticism about the practical implications of the project: “Finding new ways, as these scientists do with robots, to try to recover rare earth metals is very important. I am not convinced that at this point it can be done on a large scale“.
The idea of using AI is quite new, especially for its application. Suffice it to say that Apple has attempted something similar, but of course only considering iPhones. A “universal” solution it could find a place in the plants of any recycler and be used for many different models. The project wants to demonstrate at least the feasibility of the idea.
Each of the teams will have different tasks: INL researchers will develop methods for removing batteries from smartphones with a robotic arm; those of the University of Buffalo and Iowa State University to identify valuable components (cameras, magnets etc.) removable by robots and to find or develop a robot capable of doing so. The software will be crucial to the success of the project and will be researchers from Iowa State University and Sunnking a create a database with 2D images and 3D scans of various smartphone brands and models.
Through machine learningthen, the database will be used to train the software that will guide the robot in its operations. In this regard, Neal Yancey of the INL stated: “We will train the system to look at smartphones and distinguish an iPhone from a “Samsung XYZ model”, and then turn to the database to understand how to operate to remove the battery“.
Sunnking will provide 100 samples of five different smartphone models to researchers to conduct experiments and will then be the first recycler to test the system towards the final stages of the two-year project.
The INL researchers will also be entrusted with the task of analyzing theeconomic aspectto understand if a robot with AI can actually reduce the costs of recycling. The stated goal is to recover at least 10 percent more materials and at the same time reduce current costs by at least 15 percent. Although these numbers do not seem revolutionary, the project is already complex – as pointed out by Minghui Zheng of the University of Buffalo, one of the most important figures in this project – due to the difficulty of finding a robotic arm suitable for carrying out non-simple tasks in quickly and accurately.
Possible difficulties of the project and the fault of the producers
They already exist now projects who use AI to work with solid plastic waste; other groups such as Carnegie Mellon University’s Biorobotics Lab (which recently partnered with Apple) are trying to develop artificial intelligence-based e-waste recycling methods.
Although the idea behind the project is interesting and certainly full of potential, his practical use it is still quite distant: just the fact of starting from five models in all (against the thousands in circulation) offers the idea of the amount of work to be done; moreover, the system will have to be scaled to work large volumes of smartphones in an industrial context.
All this without forgetting that the design of smartphones evolves over time, therefore the robots will have to be continuously updated at the hardware and software level to keep up. A recycler inclined to invest in this technology should also take into account the fact that a purchased robot – these are complex jobs, therefore there is still no talk of possibility of renting – could become obsolete within a decade.
According to Sara Behdad – researcher in the field of sustainable electronics at the University of Florida, not involved in the project in question -, the real difference will have to be made by smartphone manufacturers, by designing devices that are easier to recycle: Robots may be useful, but many current problems are caused directly by manufacturers, for example by gluing batteries and using proprietary screws. For Behdad it would be essential to address these problems through the adoption of standards for disassembling smartphones.
Read also: Where do I throw my old smartphone? How to recycle electronic devices