Game Design, Programming and running a one-man games business…

My non games idea

I’d like to set up a company one day which solved an inefficiency in the market for consumer services.

Right now, in the UK and most countries individual consumers enter into deals with utility suppliers to provide them with Electricity, maybe gas, water, and telephony / internet access. This involves setting up an account, deciding on a username/password, some bank details, picking a ‘plan’ and so on. We are widely told that the range of deals available means we should shop around and change our providers often. In practice few people do this. (I do, but I’m…different). What bugs me is not that few people do shop around, but that the process is so clunky and manual.

Now I’m not talking about ‘making switching easier’. thats lame, and unambitious. I’m talking about making it automated. Not on a year-to-year basis, but on a second by second basis. Think less like a form-filling bureaucrat, and more like a high frequency trader.


When I switch on my PC, and it draws current from the mains, I want there to be a super-fast auction, right then, in the exact same way banner-adverts are traded, where my AI agent that represents me haggles on the energy market to get me the extra power for the next few seconds. And I want it to keep haggling all the time I’m drawing power. I also want my taps (faucets to some of you) to do the same thing, and I want my telephone, broadband, everything to work the same way.

Of course, this works WAY better when we can defer demand. A smart fridge can, for example put in a  request for power to its compressor *some time in the next five minutes*, but not care exactly when. As a result, it should get a darned good deal. On the other hand, a hairdryer needs power RIGHT NOW, and at the other end of the spectrum, my fancy-pants electric car can charge *at some point in the next eight hours*. I don’t care when.

This would make for a huge boost in energy efficiency. The demand curve of UK power would flatten out substantially, meaning less need for overcapacity to handle ‘spikes’. It used to be the case that the definition of a UK power spike was either the adverts at the end of the TV show ‘coronation street’ or the end of the queens speech on Christmas day. Why? Because thats when about 10 million Brits switched on a 240volt 2,000 watt electric kettle. I’m not kidding. Apparently after the ‘who shot Phil Mitchell’ episode of EastEnders, power surged by 2.6 GW. As I type this, demand is 32GW in total.


Anyway…what annoys me is not that we do not have this system in place (I understand it involves huge infrastructure investment and new appliances), but that we do not seem to be making ANY steps in that direction whatsoever. I have a big energy-guzzling car, which would be perfectly suited to negotiating a time to draw current with the power company, and yet there is, AFAIK, NO provider in the UK that even has a prototype for such a system. Why?

Sadly I expect this will never happen, and what we will end up is local co-operatives handling power management themselves. if I had a Tesla powerwall, I could save any excess from my solar panels, and either use it to charge my car, or to sell into a village-linked system to a neighbour. I suspect local systems like this, with a fraction of the potential savings are going to become commonplace before any of the big players in UK infrastructure take a step in this direction.

8 thoughts on My non games idea

  1. I think this is one symptom of a problem the world i suffering from. Even in politics people seem to be most concerned with putting out fires and winning votes instead of making thought-through long term plans. A good example of such a plan is Elon Musks about the future of electric cars. Why is that? My bet is on money. There’s more money in waste anf inefficiency I’m guessing :(

  2. The power companies have overhead: running power stations, administration, profits. If you want the investment in smart power infrastructure to pay off so that it benefits power companies and consumers, the money for that has to come from somewhere. Presumably, we could decommission some power stations that are now kept on standby, but how much savings would that entail? A couple percent? Enough to refinance smart power? There’s a huge up-front investment by both the utilities and the appliance manufacturers (passed on to the consumers who buy them), and while smart appliances that can control their own power usage are not a new idea, I think this economic deadlock precludes their market introduction. Mind you, car chargers could do it, mostly because they would save the utilities from investing in new infrastructure that would become necessary if demand rises from this.

    1. Well there is slow ‘churn’ in power stations now as coal ones get switched off and new renewable ones come online, so there is *some* scope for it. I agree that car charging is the perfect opportunity to take steps in this direction, as its something that will expand, and its for BIG amounts of power, plus all of the chargers for cars have custom negotiating circuitry in them already, so not much of a leap.

  3. The electricity market works largely in the way you describe, but at a much less granular level of demand/supply (at least, in the Australian one, with which I am familiar). Getting a licence to trade in the market is currently expensive (too expensive for a fridge, or you, or your suburb).

    What is not clear is whether the evolution of markets to provide such facility is the easiest / cheapest way to solve the problem. Of course, we currently are doing very little to solve the problem. Perverse incentives for some involved do not help. Anyway, its not like there is anything important at stake…habitable planets are commonplace in some solutions to the Drake equation.

      1. While I like the idea of a dynamic scheduling for power draws, I’m not sure where the advantage of modeling it on a trading floor comes in. Power distribution is essentially a routing and allocation problem, isn’t it? The advantage to adding a reservation system like you propose – “I want x power at y time” as opposed to “I need x power now” – is that you’re relaxing the constraints on your routing solution. That’s a good thing because it means that the system can more efficiently pack allocations, leading to a higher average system utilization and less of a requirement for a safety overhead in the maximum system capability. It’s a good argument for an intelligent and dynamic power grid.
        What I don’t see the argument for is adding the “price of power” constraint to the system in the form of a trading floor. If you’re trying to engineer the system for minimum waste, you want a utility function emphasizing minimum line-loss and a minimum amount of generating resources active. Most power generation doesn’t have meaningful output controls beyond “is running/is not running” and some don’t even have that – there’s literally no benefit to not using all of the power produced by a solar array, for example. The main effect of adding a bidding floor would be to add inefficiency to the “bring extra generation on line to meet current demand” part of the process. If you have reserve coal plants, for efficiency you want to spin them up one at a time based on their location in the distribution graph, not based on the price-point their respective operators treat as their “it’s worth selling power now” point.

        1. interesting.
          I guess what I’m thinking is that power demand can also have price points based on time. For example my car needs power before 8AM, but if its cheap and its 3AM, I’ll have it now. If I still haven’t got any and its 7AM, I’ll pay more for it. Price used purely as a means for regulating non time-critical demand.

  4. Hi Cliffski,

    The powerdraw to charge your Tesla overnight…

    Isn’t that was used to be called Economy 7 (useless electric home heating from the 70’s / 80’s…) but it was there because it was impossible to “switch off” oil or coal burning powerstations (which still provide most of the power to the UK) so it was a way of “storing” power that otherwise would have been wasted.

    I have to say I take a different view…

    Whereas it is possible to storage gas (and therefore buy it at cheap prices) or simply not dig coal out of the ground depending on its price in the market when these are used for heating or cooking then you can “manage” demand but both of these are still fossil fuels and shouldn’t be used (unless in an emergency) and I still think solar has a way to go if you looked at generated power per Sq M / Per year vs bio fuels or even food…

    Unfortunately the demand for electric power will always be just that – a demand driven requirement – would it be possible to plot the requirement for lightbulbs or have them “store” electric based on a 8 hour power cycle if so you say sun set is 17:03 PM – Want to tell that to my kids ?

    I have to say that electric power in the future should be generated by nuclear power and to keep it safe and affordable it should be provided by the government as a general national infrastructure whereby you look at the cost model (build / running over a 30 year period / decommissioning (is it bad to take spent fuel rods and put them in a canister and fire them in to the Sun on an old last mission rocket?) and how much power the station can provide over its life span which would give you a pence per MegaWatt ?

    The current problem is we have sold out to big power companies that now control the manufacture and distribution (along with Water, Sewerage and Treatment and shortly Healthcare) of things that the whole collective of the UK have invested in over the last 100 years or so and now can hold us to ransom as these are things we all need and have no where else to go to purchase….

Comments are currently closed.