4 ms·

Depends on the game a bit but I too am a bit sceptical that this adds much, would be interesting to see a breakdown of solar vs buttons.

by icoder 4y ago

vanderZwan 4y ago
Even if it turns out to be useless, this was a research project. Testing if this is feasible in practice is part of the point. Although it is a bit odd that they don't show any stats on the main page. I looked it up in the paper[0], section 5.2 discusses power consumption/generation and pasted it below (I hope all symbols survived the clipboard). TL;DR: the buttons generate a very small difference in power generation relative to solar panels. HOWEVER, this still seems to be a significant difference in how quickly the battery drains. If solar panels generate 10.5 mW, and the buttons 0.5 mW, but the system requires 11.5 mW, then this doubles the up-time of the system. That would actually be quit a nice insight to share on the main project page IMO! > We have measured ENGAGE’s power consumption, looking into overall power consumption whilst first measuring the consumption of MCU together with the FRAM and display. The MCU and FRAM combined consume 11.15 mW and the screen consumes 344.31 μW during game execution taking a ten second average. During idle time the screen only consumes 3.90 μW, resulting in a combined system average power consumption of 11.50 mW. As a comparison, the original Nintendo Game Boy consumes 232.08 mW during game execution, varying slightly per game and cartridge architecture. While not necessarily a useful or meaningful number, we conclude that our platform is more than 20 times more power-efficient than the original Nintendo Game Boy (representing normal technology advancement, but noting that ENGAGE is an emulator). The measurements were conducted using a Fluke 87V [33] multimeter and the X-NUCLEO-LPM01A [123] programmable power supply source with power consumption measurement capability > Then, to give more insight in the energy harvesting on the ENGAGE platform we have measured the amount of energy the solar panels generate using a Fluke 87V [ 33 ] multimeter and compare this to the energy generated from the buttons. For the buttons we use the minimal energy generation figures from the specification of the harvester [ 153 , summary] as a worst case scenario3. Assuming that a single button press generates a minimum of 0.66 mJ and knowing the amount of button presses per game is specific to the game as per Table 1 , we can assume the buttons generate between 0.66 mJ for one press per second and 1.98 mJ for three presses per second. At 40 klx and 20 klx, the solar panels generated an average of 10.14 mW and 8.33 mW, respectively, i.e. less than the required system average power consumption of 11.50 mW. We can conclude that ENGAGE is mostly powered by solar panels and supplemented by the button presses although the button presses can significantly increase the on-time of the platform, as shown in Section 5.1 [0] https://dl.acm.org/doi/pdf/10.1145/3411839 https://dl.acm.org/doi/pdf/10.1145/3411839
- ajsnigrutin 4y ago
  But the idea of a research project is to find out what is worth doing and what isn't. Do shitty, hard to press buttons actually help? Or would putting just a tiny bit larger solar cells help a lot more. If you're in complete darkness, can you actually press enough buttons to make it playable there too? There have been many, many really shitty solar projects in the last year, some wasting millions of dollars and euros, and while reasearch is important, it's also important to know what's worth it and what's not.
  - vanderZwan 4y ago
    So the questions in your first paragraph are mostly answered by the paper. And your second paragraph feels a bit like an unfair comparison to me. This is a first attempt at seeing if this concept is even remotely feasible, and it uses off-the-shelf parts for that. It's not a company with a lot of money and resources invested in it that produces mediocre final products where the embodied energy of the solar panels is so much more than any energy it saves over its life-time that it's less green than something that runs on batteries. Which, yes, are an annoying feature of modern greenwashing. But this research project is not that.
    - ajsnigrutin 4y ago
      But you'd expect a research project to say "we tried harvesting button presses, rf, inertial energy, and all of it sucked except for solar, so we decided to up the solar as much as we can with a gameboy sized case". If they have budget issues, then buy one of those energy harvesting buttons, press it as fast as you can and measure the output... if it generates any meaningful amount of power, nice, buy one for every button and use it in the final "product"... if not, try something else. Solar cells are cheap, 3d printing is cheap, so adding a few more would help a lot.
      - vanderZwan 4y ago
        > But you'd expect a research project to say "we tried harvesting button presses, rf, inertial energy, and all of it sucked except for solar, so we decided to up the solar as much as we can with a gameboy sized case". Why do you assume this? It's academic research, not the R&D department at a company. They're not building a product to sell and they're probably working with a shoestring budget. The goal is to extract as much knowledge and insight as possible from as little hardware as possible.

Depends on the game a bit but I too am a bit sceptical that this adds much, would be interesting to see a breakdown of solar vs buttons.

by icoder 4y ago

Depends on the game a bit but I too am a bit sceptical that this adds much, would be interesting to see a breakdown of solar vs buttons.

vanderZwan 4y ago

Even if it turns out to be useless, this was a research project. Testing if this is feasible in practice is part of the point. Although it is a bit odd that they don't show any stats on the main page. I looked it up in the paper[0], section 5.2 discusses power consumption/generation and pasted it below (I hope all symbols survived the clipboard). TL;DR: the buttons generate a very small difference in power generation relative to solar panels. HOWEVER, this still seems to be a significant difference in how quickly the battery drains. If solar panels generate 10.5 mW, and the buttons 0.5 mW, but the system requires 11.5 mW, then this doubles the up-time of the system. That would actually be quit a nice insight to share on the main project page IMO! > We have measured ENGAGE’s power consumption, looking into overall power consumption whilst first measuring the consumption of MCU together with the FRAM and display. The MCU and FRAM combined consume 11.15 mW and the screen consumes 344.31 μW during game execution taking a ten second average. During idle time the screen only consumes 3.90 μW, resulting in a combined system average power consumption of 11.50 mW. As a comparison, the original Nintendo Game Boy consumes 232.08 mW during game execution, varying slightly per game and cartridge architecture. While not necessarily a useful or meaningful number, we conclude that our platform is more than 20 times more power-efficient than the original Nintendo Game Boy (representing normal technology advancement, but noting that ENGAGE is an emulator). The measurements were conducted using a Fluke 87V [33] multimeter and the X-NUCLEO-LPM01A [123] programmable power supply source with power consumption measurement capability > Then, to give more insight in the energy harvesting on the ENGAGE platform we have measured the amount of energy the solar panels generate using a Fluke 87V [ 33 ] multimeter and compare this to the energy generated from the buttons. For the buttons we use the minimal energy generation figures from the specification of the harvester [ 153 , summary] as a worst case scenario3. Assuming that a single button press generates a minimum of 0.66 mJ and knowing the amount of button presses per game is specific to the game as per Table 1 , we can assume the buttons generate between 0.66 mJ for one press per second and 1.98 mJ for three presses per second. At 40 klx and 20 klx, the solar panels generated an average of 10.14 mW and 8.33 mW, respectively, i.e. less than the required system average power consumption of 11.50 mW. We can conclude that ENGAGE is mostly powered by solar panels and supplemented by the button presses although the button presses can significantly increase the on-time of the platform, as shown in Section 5.1 [0] https://dl.acm.org/doi/pdf/10.1145/3411839 https://dl.acm.org/doi/pdf/10.1145/3411839

ajsnigrutin 4y ago
But the idea of a research project is to find out what is worth doing and what isn't. Do shitty, hard to press buttons actually help? Or would putting just a tiny bit larger solar cells help a lot more. If you're in complete darkness, can you actually press enough buttons to make it playable there too? There have been many, many really shitty solar projects in the last year, some wasting millions of dollars and euros, and while reasearch is important, it's also important to know what's worth it and what's not.
- vanderZwan 4y ago
  So the questions in your first paragraph are mostly answered by the paper. And your second paragraph feels a bit like an unfair comparison to me. This is a first attempt at seeing if this concept is even remotely feasible, and it uses off-the-shelf parts for that. It's not a company with a lot of money and resources invested in it that produces mediocre final products where the embodied energy of the solar panels is so much more than any energy it saves over its life-time that it's less green than something that runs on batteries. Which, yes, are an annoying feature of modern greenwashing. But this research project is not that.
  - ajsnigrutin 4y ago
    But you'd expect a research project to say "we tried harvesting button presses, rf, inertial energy, and all of it sucked except for solar, so we decided to up the solar as much as we can with a gameboy sized case". If they have budget issues, then buy one of those energy harvesting buttons, press it as fast as you can and measure the output... if it generates any meaningful amount of power, nice, buy one for every button and use it in the final "product"... if not, try something else. Solar cells are cheap, 3d printing is cheap, so adding a few more would help a lot.
    - vanderZwan 4y ago
      > But you'd expect a research project to say "we tried harvesting button presses, rf, inertial energy, and all of it sucked except for solar, so we decided to up the solar as much as we can with a gameboy sized case". Why do you assume this? It's academic research, not the R&D department at a company. They're not building a product to sell and they're probably working with a shoestring budget. The goal is to extract as much knowledge and insight as possible from as little hardware as possible.