This page is ded­ic­ated to present some ideas which come to my mind. This can be everything, even bad ideas…

Food re­lated ideas

  • Auto­mat­ic food re­cipes trans­la­tion, in­ter­na­tion­al web­site: When my wife searches for a re­cipe she looks at a French web­site, as this is her mother lan­guage. And a French web­site mostly has re­cipes which are com­mon in French. I do not ex­pect that someone can find a typ­ic­al re­cipe from Thail­and there. The wife of a friend is Thai, and she cooks very de­li­cious Thai meals. The in­gredi­ents are sim­il­ar to stuff we can find here, pork meat and spices and so on, but you have to search around to find such re­cipes on a Ger­man or French or whatever food re­cipes site. May­be you are even un­lucky and can not find a re­cipe in a lan­guage you un­der­stand. So my idea is sim­ple. In­stead of writ­ing down the re­cipes as text on a web­site, an in­ter­na­tion­al web­site with the food re­cipes of the world could work with ob­jects and pro­ced­ures. If there is no ob­ject (in­gredi­ent), cre­ate one and the cre­at­or can provide the trans­la­tion at least for those lan­guages he knows the trans­la­tion for (or only for his mother lan­guage). People which want to trans­late stuff should get a way to ask for un­trans­lated ob­jects in lan­guage X in the web­site in­ter­face. For the pro­ced­ures it is also easy. If there is no pro­ced­ure for what you want to do, cre­ate a new one and write the text in your lan­guage. Pro­ced­ures can have mul­tiple in­puts and out­puts. Ideally a pro­ced­ure is broken down in­to the smal­lest com­pon­ents (“mix A and B, res­ult C”, “use ob­ject D on C, res­ult 5*E”). A pro­ced­ure can be split in­to syn­tax (in­put type, num­ber of items per in­put type, in­put unit, op­er­a­tion, res­ult), and se­mantic (the words used to de­scribe the pro­ced­ure). An ex­ample would be “Take %%ITEM1_NUM1%%%ITEM1_UNIT%% %%ITEM1_​TYPE%% and put them in­to %%ITEM2_NUM%%ITEM2_UNIT%% %%ITEM2_​TYPE%%. This is now %%RESULT1_NUM%%%%RESULT1_UNIT%% %%RESULT1_​TYPE%%.”. With 3 Eggs and 1 bowl as in­put this would be “Take 3 Eggs and put them in­to the bowl. This is now the bowl with eggs.” (the soft­ware is re­spons­ible to pick the cor­rect trans­la­tion of type de­pend­ing on the num­ber). An­other pro­ced­ure could be to scramble what is in “bowl with eggs”. Yes, in case of type “bowl with eggs” this can cre­ate a lot of trans­la­tion ef­fort needed be­cause of the huge pos­sib­il­it­ies what is in­side, but may­be a “bowl with in­gredi­ents (step X)” is user­friendly enough. Cre­at­ing such a re­cipe is more work than just writ­ing down in your mother lan­guage what needs to be done. The be­ne­fit is on the con­sumer side. People look­ing up a re­cipe could chose from a huge num­ber of world­wide re­cipes (if this ser­vice would get enough con­tent pro­du­cers and trans­lat­ors), and the pro­ced­ures could even have links of de­tailed de­scrip­tions how to do some­thing, or links to videos/​animations/​pictures show­ing how to do a pro­ced­ure. For cul­tures where are spe­cific in­gredi­ent is not al­lowed for re­li­gious or whatever reas­ons, there could be an opt-​in for such in­gredi­ents (de­tec­tion based upon browser lan­guage and/​or geo­graph­ic­al re­gion).
  • No-​name products track­ing: Write an ap­plic­a­tion for mo­bile phones (if pos­sible even for sim­ple ones which come only with Java, e.g. via in­tern­al DB lockup in­stead of online-​lookup, not only for An­droid/​iPhone) which scans the pro­duct bar­code, the date (of last con­sump­tion or man­u­fac­tur­ing date or whatever), and the num­ber of the man­u­fac­tur­ing com­pany (at least man­dat­ory in the EU for meat, fish, milk, eggs and sim­il­ar) and re­turn the name(s) of sim­il­ar no-​name or even branded products. With the num­ber of the man­u­fac­tur­ing com­pany you can see which brand could be the pro­du­cer of the pro­duct. While this does not mean it is the same (in­gredi­ents and/​or qual­ity), so does it give a hint what to buy or to avoid for whatever reas­ons. With the pro­duct bar­code and the date you could get the pro­duct his­tory (when the pro­du­cer changes). With user sup­plied in­put to the main data­base the match can be im­proved (e.g. “tastes the same as the branded one”). It also al­lows to find cheap­er al­tern­at­ives to branded products, or to see if some­thing sim­il­ar is also avail­able at an­other place. Ad­di­tion­al user sup­plied in­put (or may­be even of­fi­cial in­put from com­pan­ies) could be the (re­gion­al) price. This would al­low to find the cheapest price for this kind of pro­duct.

Al­tern­at­ive en­ergy re­lated ideas

  • Non-​optimal weather op­tim­ized photo­vol­ta­ic sol­ar cells: Could the av­er­age power gen­er­a­tion of a photo­vol­ta­ic sol­ar cell in non-​full-​sunlight con­di­tions be im­proved by us­ing a fres­nel lens (the Wiki­pe­dia art­icle sug­gests that it can be used there, but so far I have not seen any use of it on small (private house­hold) in­stall­a­tions)? What im­pact would this have on the peak power gen­er­a­tion? The same ques­tions for thermal sol­ar cells. UPDATE 2010-​03-​16: I found a com­pany which seems to be do­ing this. They use a heat-​exchanger to move the ad­di­tion­al heat as dis­cussed in the com­ments here to a wa­ter pipe to gen­er­ate hot wa­ter (a com­bin­a­tion of a photo­vol­ta­ic and a thermal sol­ar cell). They use the lens and some mir­rors to re­duce the amount of sol­ar cells (and thus the cost). Un­for­tu­nately they do not com­pare the out­put of the photo­vol­ta­ic part with a nor­mal sol­ar cell (the same sur­face area but no lens), they only com­pare the com­bined out­put power of the photo­vol­ta­ic and thermal parts with a nor­mal photo­vol­ta­ic pan­el (so they com­pare apples with or­anges and show a huge dif­fer­ence in this un­fair com­par­is­on).
  • Lo­tus flower ef­fect for sol­ar (photovoltaic/​thermal) cells: If the glass of a sol­ar cell would be mod­i­fied to show a clean­ing ef­fect of the lo­tus flower, would this re­duce the amount of main­ten­ance? If yes, is this sig­ni­fic­ant? Would this ef­fect cause a change in the ef­fi­ciency of the cell? If yes, how much?
  • Us­ing com­pressed air to auto­mat­ic­ally re­move snow from sol­ar cells: Is there an en­ergy ef­fi­cient way to auto­mat­ic­ally clean e.g. snow with com­pressed air from sol­ar pan­els when the time of the day sug­gests that there could be sun but a light sensor does not de­tect more light than dur­ing the night (this “the pan­els are covered by some­thing” de­tec­tion does not take in­to ac­count that pan­els can be par­tially covered, and may­be this/​it could be im­proved by tak­ing the power out­put of the pan­els in­to ac­count too)? The power con­sump­tion of the light sensor and the con­trol elec­tron­ics need to be taken in­to ac­count. Is this a sig­ni­fic­ant im­prove­ment (with a pos­it­ive re­turn of in­vest­ment)?
  • Gen­er­at­ing elec­tri­city with heat: Thermal sol­ar cells gen­er­ate more heat power than sim­il­ar sized photo­vol­ta­ic cells pro­duce elec­tric power. How ef­fi­cient would it be to use thermal sol­ar cells to drive e.g. a stirl­ing en­gine to gen­er­ate elec­tri­city in a nor­mal house­hold (I have only found big in­stall­a­tions, but no in­stall­a­tion for a nor­mal house­hold)? To cool down the rel­ev­ant part of this setup, the ex­cess­ive heat could be used to heat the house or the wa­ter for do­mest­ic use.

Car re­lated ideas

  • Constant-​power cruise con­trol: Now with more and more elec­tric cars be­ing an­nounced, this would al­low to cre­ate a cruise con­trol which can be switched from con­stant speed to con­stant power. The con­stant speed mode is like we all know how a cruise con­trol op­er­ates (de­signed to bet­ter cal­cu­late the ar­rival time… in the be­st case), the con­stant power mode would have a vary­ing speed, but would try to have a con­stant power out­put per distance-​unit (de­signed to bet­ter cal­cu­late the travel dis­tance the power source is able to handle).
  • Per­son­al GPS track­ing: It would be nice if the GPS devices in cars would grow a track­ing fea­ture which re­cords where I was, how was I drove, and at which time I was there, and at which day of the week (opt-​in for each time, off course). With this it would be pos­sible to see for ex­ample if it mat­ters if I drive 20 speed units faster on a spe­cific way at a spe­cific time or not (not dir­ectly, you would have to drive of­ten the same way to get some stat­ist­ic­ally sig­ni­fic­ant an­swer). It would also al­low to gain bet­ter in­sight in­to which way to take at which time of day to get from e.g. work to home.

2 thoughts on “Ideas”

  1. Us­ing a fres­nel lens to fo­cus a lar­ger quant­ity of light on­to a smal­ler area is good, how­ever keep in mind that cur­rent photo­vol­ta­ic sol­ar cells are 10 – 20% ef­fi­cient only. The idea is good, but there is more to it than simply aim­ing a fres­nel lens at it, since dur­ing peak times you prob­ably do not want the fres­nel lens fo­cus­ing that en­ergy down on to the smal­ler area (it will prob­ably make the sol­ar cell less ef­fi­cient dur­ing those peak times).
    I am glad to see you think­ing about al­tern­at­ive en­ergy!

  2. Thanks for the in­fo. Some in­ter­est­ing in­fo would now be, if the av­er­age power gen­er­a­tion (photo­vol­ta­ic) is higher or not then (may­be with some ad­di­tion­al (non-​electrical) cool­ing im­prove­ments), and how much a fres­nel lens can cost from an eco­nom­ic point of view. But this highly de­pends on the in­stall­a­tion en­vir­on­ment.

    Re­gard­ing al­tern­at­ive power gen­er­a­tion… I even have some links to DIY thermal col­lect­ors which look in­ter­est­ing, but without heavy changes in my house first (adding just the tubes from the heat­ing room to the roof means that I have to ren­ov­ate the bath as I have to des­troy a good part of it), I can not ex­plore this in­ter­est­ing field.

Leave a Reply

Your email address will not be published. Required fields are marked *