Sunday, 23 December 2012

Social network cafe

Whilst smartphones (and smaller tablets) are getting better for checking social networks on the go, there is as yet no substitute for the larger screen real estate of laptops, desktops and larger tablets. As such their is an opportunity for social networks to create cafes where people come and can use a laptop/desktop/tablet to use the social network. The laptops/desktops/tablets would need to be tethered to the tables to prevent theft, and would need robust security to prevent account hacking.

The advantage to the consumer would be free access to larger screen real-estate computers on the go. The advantage to the social network would be to gather significant amounts of real-world user-experience data: mouse and keyboard logging; video of the user's activity; etc. Obviously this snooping would need to be disclosed to the cafe users, but it would not be an unfair trade-off. Particularly if the price of coffee was right!

Congestion-based road pricing

Roads are congested different amounts at different times. This indicates that the disincentive of sitting in traffic is insufficient to dissuade drivers from driving. So congestion charging makes sense.

But most schemes (at least those I've read about), are fairly static in nature: they don't take dynamic day-by-day (or minute-by-minute) congestion data into account when establishing zones or charges.

A system that did take account of dynamic data is not beyond our technological capabilities. A congestion fee per road, per direction on that road, per hour of the day, per day of the week and per time of the year could be established dynamically from traffic-monitoring data and applied to a driver's account. Drivers would be able to view their accounts and see the journeys that were contributing most to their congestion charge, and they may then choose to alter their journey habits. They could review available pricing data to identify cheaper periods. The market forces at work, operating at a micro level (per road, per minute) would allow for much more efficient use of our road network.

Such a system would also work in our driver-less car future. The fee for using a shared driver-less car fleet could incorporate a congestion component, again at a per road, per direction on that road, per hour of the day, per day of the week and per time of the year, level of detail. This would again use market forces to encourage behaviour that maximises the efficient use of our road network.

Monday, 17 December 2012

Wikileaks for prices

With notable exceptions marked "price on application" (which are really only for rich people anyway), pricing data in the business-to-consumer market is abundant. Price comparison websites have made the job of accessing this already-available price data all the easier.

But in the business-to-business market, deals are still done behind closed doors, and pricing data is more difficult to come by. What we need is a Wikileaks for B2B pricing. Participants could anonymously submit pricing data about goods and services that they had bought, making for a more competitive and dynamic market.

(This is really just a more specific version of my suggestion of a heuristics website).

Boldy going

Will the human race or our descendents explore the universe? Or will we spend our time exploring the richer virtual universe, whilst our robotic slaves harness the physical universe to create more computational power to make our virtual universe all the richer?

What to wear

I'd like an app (ideally on my phone) that would allow me to record what I was wearing and whether appropriate for the conditions. This personal assessment of attire versus conditions could be mapped to weather data as derived from sensors on the phone, or via location-sensing and download of weather data.

The result of this data collection would be that the app would be able to recommends what to wear on a given day based on the weather forecast. And it avoids me having to consciously map closing to weather (answering the "yes/no" appropriateness of current attire question is much easier). In particular it helps with the hard-to-assess areas such as the impact of wind chill factor depending on layer combinations.

Elevated baby playground

Why do parents chose to play with children down at their level, rather than raising them up to a suitable height? Probably the answer to this question is safety. But anyway, why not build a set of raised platforms that babies/toddlers can crawl around, with guard rails to stop the babies falling off the sides. The platforms could be arranged into a ring, with a lifting bridge allowing adults to enter the centre of the ring. The base of platforms could be used for storage, making efficient use of space.

Knowledge and creativity

Knowledge helps you access as-yet unploughed fields of creativity.

Sunday, 16 December 2012

The ultimate acquisition

High synergy, low market covariance

Note folder

This thought is in some ways (or all ways) entirely pointless as paper notebooks will soon be superseded by tablets. But nonetheless, here goes:

A common problem I encounter with my paper notepad is that it tends to be filled chronologically. Whilst this is useful for somethings, it does not allow grouping of notes by topic.

An alternative approach that does allow notes to be re-ordered chronologically would be loose-leaf sheets. This is in fact how I took notes back in college, with those sheets being added into the relevant lever-arch file. But such an approach is not common in a professional setting despite its benefits.

There are a couple of reasons for this: firstly, the volume of notes professionals take is typically less than students, so level-arch files are too large; and secondly, there is no professional-looking stationary for performing such a function.

The latter point suggests a opportunity, to create a paper tool that combines a pad of loose-leaf sheets with a tabbed organiser that the pages and be quickly and easily added into. This could be for example a (slightly larger than) A3 side leather folder with a seam down the middle. The right side (for right-handed people) holds a pad of paper, the left holds a ring-binder with tabs to drop the pages into.

Monday, 3 December 2012

Solar 3D printed bricks

I've mentioned previously about Markus Kayser's SolarSinter, which uses the energy of sunlight to fuse sand into three-dimensional shapes. I've been thinking since then about the ability to create the SolarSinter technology to create bricks in an automated fashion.

Today I came across Building Bytes, who are experimenting with 3D printed bricks that slot together, in a similar fashion to LEGO.

By combining these two projects, plus a bit of solar-powered automation, it should be possible to create a device that takes in desert sand at one end, filters it, sinters it and outputs bricks at the other. Bricks that do not require cement, and are easily stacked into buildings.

Taking it a step further, the GRASP Lab has created quadcopters that can take autonomous components and construct a complex structure. So, sand goes in one end, and buildings come out the other. With little human supervision.

Sunday, 2 December 2012

Product will no longer be made notification

The IT sector is reasonably good at letting its customers know that products will go out of production, and out of support. But for the wider manufacturing sector this information is not reliably communicated to end users.

One benefit of unified online datasets will be the opportunity of users to enable this information to be passed to them in the form of notifications. For example, the dataset might be a store card that logs a user's purchases. The user can then opt to be notified if any of their regular purchases, or a component part for one of their purchases, is going out of production. The user can then chose to stockpile, or potentially petition the manufacturer to keep the product in service. If enough users do the latter, the product may stay in production.

Fish eye glasses

Continuing today's optic theme, I've been thinking about the possibility of using the concept of fish eye lenses to give humans a wider field of view: fisheye glasses.

Such glasses could probably be achieved with a set of convex and concave mirrors: the convex mirror would go on the forehead capturing light from a wide field of view, reflecting it into a concave mirror that would sit in front of the eyes and reflect the wider field of view into the eyes.

Obviously the field of view would be distorted compared to what we are used to viewing, but humans are remarkably adaptable, and would soon become acustomed to the new field of view.

Wiring up datacentres with lasers and mirrors

Would it be possible to replace the copper and fibre optic wiring that permeates datacentres with a system of lasers and mirrors? The mirrors would be actuated, allowing the connectivity within the datacentre to be adjusted with minimal effort. Obviously, like fibre optics, the communication would be at the speed of light.

It should be possible to have mirrors configured to allow beams to "jump over" other beams, and thus prevent any beams crossing.

Unlike wireless (e.g. wifi) solutions, there is no shared medium, and hence no interference.

Sunday, 18 November 2012

Deeper down the rabbit hole

The physical world is only there to support our excursions into the virtual. We will reach a tipping point, in some aspects we already have, where the virtual world will provide experiences more vivid than the physical.

Sunday, 11 November 2012

Pushing yourself

From time to time seek to push your limits in a controled environment; generally operate at your performance optimum.

The purpose of humanity

Creation: creation of life; creation of ideas; creation of knowledge; creation of culture.

Wednesday, 7 November 2012

One guaranteed prediction for the future

The future will be less interesting than I want it to be.

Eventually open source licence

Would it be possible to write a legally valid eventually open source licence? Such a licence would include a commited date at the point of publication of when the code would be released open source. It could potentially require users of the licence to submit source code into escrow to be released at that date.

The licence would balance the commercial needs of making return on investment, whilst making a reputation-maintaining commitment to the community, and also future-proofing such that users software is not locked into unsupported closed-source formats.

Expanding the MMORPG grinding concept to websites

In MMORPGs, players engage in repetitive tasks to advance levels or get money etc (a process called grinding). There is potential to expand this model to websites to drive user engagement:

1. Users earn credits by being active on the site (posting comments, messaging friends, reviewing content, etc).
2. Credits can be used to do certain actions that are not possible without them (functionality that is desirable to users).
3. Credits can be sold to other site users for real money.

Time rich cash poor people can earn money by generating excess credits from their engagement, which can then be sold to time poor cash rich people who want to access the locked features.

The features would not be purchaseable other than via the credits bought from other users. This mechanism would create a supply/demand balance that should drive user engagement with the site.

How does the site make money? Advertising obviously, but also the site takes cut of credit to cash trades.

Some examples of activities on websites that could generate credits:
  • connecting people - put two people together who both subsequently confirm friendship
  • sharing content that friends "like"
  • encouraging their friends to visit the site
  • acting as a moderator or admin
  • making comments that use positive language (might need an irony detection algorithm...)

Judging time

Observation: we think about time differently if we see it on a clock face or presented in digital format.

Experiment: assess how people make time based decisions (e.g. how fast they work to get a task done within 10 minutes) whilst varying the format of time presentation (analogue/digital/etc)

Components of the smart home

Below is the first draft of a table of components of the smart home. The key is that all of these components will be capturing data and/or actuating the environment. That data will be centrally stored within the house computer, and intelligently linked. A good example might be windows opening to generate airflow based on a combination of temperature and wind direction data, combined with the temperature preferences of the various people within the homes and the various rooms they are in.

Category Component Sensors Actuators Data stored / logged Uses
Building integrity sensors Structural sensors

Building integrity sensors Moisture / water sensors

To detect leaks in the roof and plumbing; or rising groundwater/flooding
Communications Phone

Communications Femtocell

Communications Broadband router

Logging Internet use
Energy Solar cells

Energy Wind turbines

Energy Hot tank Level
Level over time
Energy Geothermal

Energy Deep water source cooling system (pond/lake)

Energy Backup generator

Energy Energy store (e.g. battery)

Human sensory input Screens

Human sensory input Speakers

Human sensory input TV

Logging TV user over time per channel; energy use
Human sensory input Radio

Logging radio use
Human sensory input Deodorant / air freshener

Human sensory input Wallpaper (eInk)

Environmental control Windows Open/Closed Open/Close Open/Closed status over time Maintaining air flow and house temperature by opening and closing the right windows based on prevailing wind direction; outside air temperature; inside air temperature; etc
Environmental control Shutters Open/Closed Open/Close Open/Closed status over time
Environmental control Blinds Open/Closed Open/Close Open/Closed status over time
Environmental control Curtains Open/Closed Open/Close Open/Closed status over time
Environmental control Doors Open/Closed Open/Close Open/Closed status over time
Environmental control Fans On/Off; Speed; Angle On/Off; Speed; Angle

Environmental control Air conditioning On/Off/Settings
Usage log; energy use
Environmental control Heaters On/Off/Settings
Usage log; energy use
Environmental control Lights On/Off/Settings
Usage log; energy use
Environmental control Humidifier On/Off/Settings
Usage log; energy use
Environmental control Dehumidifier On/Off/Settings
Usage log; energy use
Environmental sensors Air quality (pollutants, pollen, etc)

Environmental sensors Carbon monoxide detector

Environmental sensors Geiger counter Radiation
Radiation over time
Environmental sensors Thermometer

Environmental sensors Humidity sensor

Environmental sensors Barometer

Environmental sensors Fire alarm

Environmental sensors Air flow sensors

Environmental sensors Infrared sensors

Identify heat loss or gain from outside
Environmental sensors Odor sensor

Hot water and heating Boiler

Hot water and heating Radiators

Hot water and heating Solar water heater

Monitoring 3D image sensor (combined with image/video/object recognition)

Logging 3D image data; logging interpreted data Activity monitoring; Object monitoring (e.g. “Computer; where are my keys?”)
Monitoring CCTV

Records Repair logs

Repair undertaken; Repairman; Company; before images; after images
Records Plans

Records Manuals

Robots Cupboard doors

Open/Closed status over time
Robots Appliances On/Off
Energy use
Robots Elevator Current position; load Move floor Usage log
Robots Ramp

Robots Robot vacuum cleaner

Robots Garage door Open/Closed Open/Close Usage log
Security Burglar alarm

Security Intrusion detection sensors

Sensor log
Security Safe

Access log
Security Panic room

Water Taps Open/Closed/Flow
Usage log Alarm if tap left running
Water Shower

Usage log
Water Bath

Usage log
Water Ultrasonic flow monitor Flow
Water use

And some specific ones for the smart garden (although many of the above environmental sensors would be relevant for the garden):

Category Component Sensors Actuators Data stored / logged Uses

Soil moisture sensor

Sprinkle Usage log

Sunlight sensor

I'll cover that later...

One of the challenges of teaching is to explain a complex set of interrelated topics in a stepwise fashion. Often it is necessary to make reference to a concept that will be covered later, but without fully explaining it. Whilst this shorthand is useful, it is often not fulfilled in a rigorous fashion.

My suggestion is that teachers create list of all the things that they've said they'll cover off later. The list would be published to pupils and ticked off as covered. Each item on the list would be dated, and the context given. This would allow students to move backwards and forwards more easily in the syllabus.

Sleeper cars

I've been thinking about how the impending self-driving car revolution will change the world (and the likelihood is that it will change the world in profound ways). One in particular that occured to me was about how we use our road network: packed to a standstill at times, whilst nigh empty at others.

Self-driving cars provide an opportunity to make better use of roads at night. Already much haulage occurs overnight, but little passenger transport. This could change with sleeper cars - self-driving cars fully fitted with beds. A self-driving sleeper car of a personal rapid transport style fleet would arrive at your home the night before you needed to be somewhere. You'd get in, bed down for the night, and wake up at your destination.

The efficiency is twofold: firstly using roads at their lowest traffic load; and secondly combining travelling with sleep. (Although the latter is less relevant as self-driving cars will in general allow road travel time to be productively employed.)

Sunday, 14 October 2012


For every new thing you learn, consider how that knowledge can be assessed, both from a measuring attainment and helping learn point of view. Automated/digital assessment is incredibly valuable due to its low teacher-resource intensity. But it requires significant creativity to get right in the first place.

Innovative software

What does innovative software do? It:
  • Captures data that wasn't captured before;
  • Organises and displays data in ways it wasn't organised before; or
  • Analyses data in ways it wasn't analysed before
Capturing data that wasn't captured before includes capturing in electronic form data that was captured in another format previously (in human memory, on paper, on film, etc).

Can you think of any innovative software that doesn't derive its innovativeness from one or more of the above?

Wednesday, 26 September 2012

Assumed knowledge

The level of assumed knowledge in many situations, be they professional or social, is often unreasonably high given the diversity of such situations and hence the collective total of what would be considered assumed knowledge in at least one situation.

Why is this a problem?

Conversational awkwardness at the least. At worst, fear of asking questions leads to lack of knowledge transfer. And absence of knowledge leads to poor decision-making and mistakes.


Only engage with people with the same knowledge base. This is almost impossible to achieve, and undesirable due to the cross-fertilisation value of multi-disciplinary interaction.

Greater cultural acceptance of diverse knowledge bases, possibly driven by explicit education on the matter in the formal education system.

Pay to communicate to

How much would you pay to communicate with someone you know-of but don't know, if you could be guaranteed that they would read the message? For a fairly large proportion of the population, the answer to that question is more than zero. So there's a business model in here somewhere.

Users would pay a fee to communicate with any person (recipient) set up on the system. The price would be  dependent on market rate, i.e. more for popular people, such as to reduce the volume to a level that the recipients would cope with. The recipients signing up for the service would guarantee to read (or watch, or listen to, etc) a certain number of messages per month.

It might also be interesting to introduce the option for a (possibly partial) refund for relevant communication, i.e. messages that the recipient actually benefitted from receiving. There would likely be sufficient unwanted messages in the service that such a costs of refunds could be born, and the chance of a refund for relevant communication would also stoke demand in the user-base.

Population by age by sex

What is wrong with the above graph? In some ways, not too much, it's a very common way of presenting population by age by sex. But one of the most interesting datapoints that can be revealed by the graph is the ratio of genders in a particular age band. It is almost impossible for the human eye to discerd this as the bars are not next to each other. A more sensible approach is below. Why has such an efficient method become the standard?

Benefits of ambidexterity

Hypothesis: ambidextrous people learn more if they switch hands during writing something compared to if they use a single hand.

Test: relatively simple recall testing with a sufficient population of ambidextrous people divided into four populations (left hand no change, right hand no change, midway change from left to right, midway change from right to left)

Implication: it may be the case that there is potential value in teaching all children to be ambidextrous if the learning dividend is sufficient.

Activity time recording based on GPS data

With the data from always-on GPS devices (smartphones, in-car devices, etc) it should be possible to identify regular journeys relatively easily. A useful service based on the analysis of this data would be a monthly report on your journeys: frequency; mean journey time; median journey time; maximum journey time; minimum journey time.

It should also be possible to identify location-based activities. For example, going to the gym, going to a paritcular shop, going to work, etc. The equivalent report would include: frequency; mean time spent; median time spent; maximum time spent; minimum time spent; etc.

The general problem of being heard

In the current model of content delivery, promotion and review, mediocre pictures of cats get more attention than this blog. And let me stress, I'm not talking about good pictures of cats here, as there are some pretty amazing and entertaining ones, but the really poor ones. I'd like to thing that at least some of the content of this blog is of more value than those (although I have considered the possibility that the entire content of this blog may be of less value that a single poor quality photograph of a cat).

In the current model of delivery, promotion of valuable content is not as easy as the value of the content might suggest. In fact it's quite difficult. Marketing (promotion) to the rescue. But the problem with marketing is that it costs money, which creates a barrier. Without wanting to detract from the skill and hard work of advertisers and other promoters, I would prefer a world in which such activities were not necessary.

So how would that world work?

What's needed is more organisation and some volunteers. Each volunteer would sign up to review X amount of content per month (where X could be number of pictures, minutes of video, minutes of audio, etc), of particular genres (tag-based rather than discrete categories). Submitters would be allowed to submit Y amount of content per month, with Y increasing with successful submissions, falling with unsuccessful submissions, but never falling below a certain minimum (everyone, no matter what their past crimes, should still have a voice). Ideally (and this would of course depend on number of voluteers and the amount they volunteer to review, and the volume of content), each piece of content would be reviewed by multiple volunteers.

Additional tiers of volunteers would review anything with an aggregate positive review emerging from the tier below. The number of tiers would likely need to be higher for popular genres. Out of the top of the tier hierarchy emerges a quality content feed that non-volunteers could subscribe to (via email, RSS, etc).

What would motivate the volunteers?

Volunteers could potentially be motivated by some kind of points system for "finds" of quality content.

How does this differ from existing mechanisms?

By creating a distinct volunteer class it segregates the audience between consumers and reviewers. Although any consumer could chose to become a reviewer, the commitment that goes with it would (hopefully) instill responsibility.

A more responsible reviewer class that happened upon content that was mis-tagged would be more likely to redirect it to an appropriate reviewer, rather than that content falling ignored by the wayside.

A more responsible reviewer class would (and would be required to) give feedback to the content creator, helping them improve the creative process.

The volunteers would become experts in their niche, able to spot plagarism and also identify true original content.


No content should go unreviewed.

N.B. The irony that is post on being heard will likely not be heard is not lost on the author.

Wednesday, 12 September 2012

Spending decisions

All personal spending decisions are the replication of macro-economic resource allocation decisions on a micro scale. Every time we spend, we should be asking ourselves: is this how we want society to allocate resources? This line of thinking is potentially a useful ally in controlling our inefficient spending urges.

Carpet for stairs

Carpets are one of the most insanely unhygienic things that we introduce into our homes - they are at the bad end of gravity, and accumulate dirt, dust, stains, etc that cannot easily be removed. But they have their uses - they are comfortable underfoot and dampen noise. The latter purpose is particularly valuable on wooden stairs, which have a habit of creaking.

So what's the alternative? Small bits of rug clamped in place on each stair. These can easily be removed and washed, meeting the hygiene factor. And they can easily be replaced (to deal with staining). To ensure that sufficient grip and absence of movement on the stairs (essential to avoid injury) strong clamps are needed, but in addition to this the rug would sit on top our a rubber matting that would stand proud of the clamp.

It should be possible to design such a clamp to be both aesthetically pleasing and hold the rugs in place as tight (if not tighter) than current stair/carpet configurations.

Tractor wheel width

I've always been slightly troubled about the amount of waste associated with tractors' wheels trampling perfectly good crops - it took me a while to accept that the benefits of pesticide, herbicide and fertiliser application would outweigh the cost of trampled crops.

But there remains an opportunity to reduce the level of trampling and thus increase harvests. I've previously written about the possibility of drone helicopters applying the pesticide, herbicide and fertiliser, but I also think there is something that could be done with tractors: namely having narrower wheels.

The downside of narrower wheels is of course that they sink into the mud. But this is less of a problem for a wider diameter wheel, where the extent of sinking relative to the size of the wheel is lower. Are tractor wheel diameters at their upper maximum? I don't think so, so maybe this would work.

Basic reminder service

I previously wrote about automating memory refreshers. As a step towards seeing that idea implemented, I have built a basic reminder service.

It allows you to input topic headings, terminology, etc as you learn. And then sends a summary email of each week's submission sent a month late (or at a delay of your chosing).

Check it out.

Wednesday, 5 September 2012

Limited by shared limited by guarantee hybrid

Typically, capital is only really needed to start an organisation, for any significant changes in size and to help it in difficult financial times. Ignoring the second, and relying on debt finance for the latter, it might be possible to create a company structure that facilitates the first but allows the original shareholders to move away leaving the company in the public domain.

How might this look? The key is that the benefit that can accrue to the original investors is capped. Any profit in addition to this would be the companies (perhaps held in trust for the current and future employees of the company). As such, the original capitalist investors are rewarded adequately for the risk that they take, but eventually are able to divest, leaving the company of the employees owned by the employees for the employees (although not directly via share ownership).

Why would the original investor want to cap his/her potential earnings? Well, this could be forced by government regulation (at this point the right-of-spectrum reader has no doubt had a heart attack!), but it might be an effective way to recruit a motivated workforce. Also, the original founders might like the possibility of leaving a public legacy (surprisingly few people are purely motivated by profit).

Are there any other downsides to such an approach? The cap on the original investor's benefit results in less capital owned by that original investor allowing him/her to create new companies. However, the capital still exists and is within the hands of an established organisation to help launch new products. This could lead to an overall trend for more stable companies, but fewer innovative startups. But then again, many of the wealthy who made their money from founding new companies don't invest in innovation, so it would be very difficult to assess this either way.

Conclusion: this is worth trying out.

Sunday, 2 September 2012

What would happen if we stopped discovering new things?

The rate of change of society wouldn't slow for a long time as we are nowhere near fully exploring everything we can do with current technology, and rolling that technology out to all potential locations and uses.

Further, the technologies that enable us to make discoveries have themselves not been fully explored, so it would in fact be difficult for discovery to stop. As an example, we have the technology to sequence genomes. Each successive genome that we sequence could be argued to be a discovery in its own right, and each facilitates new discovery (e.g. finding novel useful genes).

Heating houses with distributed computing

Homes need heating, distributed computing (e.g. Folding@Home) needs doing. Servers provide an efficient form of electricity-powered heating that have the added benefit of being computationally useful.

How might such a situation work? A server-boiler would be installed much like an existing boiler, with connections to the water and electricity supplies, and additionally a network interface. Based on the home's heating requirements, the server would be timed to start up and accept distributed computing tasks. The heat from the servers is transferred to the water, and the computation continues until sufficient hot water has been generated.

From the users point of view the boiler acts entirely like a normal boiler, but with the added feel-good for them that their boiler is helping the cure diseases. The server management would be entirely remote, potentially by the distributed computing project.

Is it economical? At today's prices a server-boiler would be more expensive than a standard boiler. But as computer pricing becomes cheaper it might become cost-effective, particularly if subsidised by commercial distributed computing projects.

SaaS aggregator

As cloud-based Software-as-a-Service becomes more common, the end-user, whether consumer or business, becomes increasingly dependent on it, and hence increasingly interested in how a collection of software-as-a-service will interact. The situation is particularly acute for businesses, where businesses want all their databases to join up: CRM system; operational system; finance system; etc.

The optimum solution is a SaaS aggregator that sits between the end users and the individual SaaS services, providing:
  • shared login
  • shared databases
  • user management
  • licensing management
When logged into the SaaS aggregator, the customer can see available individual SaaS solutions that they can purchase and add into their current package. The SaaS aggregator could provide price and feature comparison, and product reviews.

The benefits of automation

How long does it take to turn off a light? Perhaps a couple of seconds, depending on how far away the switch is. How much energy does it take to turn off a light? I'm not going to guess, but it's an incredibly small amount compared with energy we consume on a daily basis.

My point is that having automated lights, or those that turn on and off at a verbal command, are not going to save us huge amounts of time: certainly less than a robot vacuum cleaner (1h per week) or a robotic car (1h per day of time that can be spent doing other things). Nor are they going to save us lots of energy (compared to say the benefit of a lift in a tall building). And yet we still want them. Why?

My suggestion is that it's not just about the time taken to do a task, or the effort expended doing a task, but that there is a cost to humans per task that is independent of time, effort, difficulty, etc. It's a kind of mental exhertion. I appreciate that this is unscientific, but that's where good science starts - with a hypothesis like this and a goal of measuring, proving or disproving.

Saturday, 1 September 2012

Fan of funnels

I was reading recently about Mass Megawatts Augmentation System, which makes use of funnels to amplify wind, similar to my post on the potential for underground wind turbines. I had originally conceived of pivoting funnels that, with the help of fins, would point towards and away from the wind. However, I wonder now whether it would be possible to have zero moving parts at the top of the device, just a set of funnels in a fan arrangement.

Depending on the wind direction, one set of funnels would get more wind than the others. Inside the body would be a series of pipes to direct the wind down, through the generator, back up and out. A system of valves could be used to ensure that the pipes connecting the active funnels (i.e. those best angled to the wind) is routed to the turbine.

Obviously this system benefits from having the turbine at ground level for easy maintenance and installation, and zero moving parts (and no maintenance) up in the air flow. Would it work?

Monday, 27 August 2012

Automating pet food delivery

The great thing about pets (perhaps with the exception of a very few fussy ones) is that they're not worried about presentation and delivery of their food, only that they get it. As such there are opportunities to automate pet food delivery without detrimenting the pets experience, whilst saving human time/labour.

It would be fairly trivial to design an automated delivery tool from scratch with the ability to alter food packaging to suite the design. Such a solution would no doubt be beneficial to food manufacturers, as it would allow them to charge a premium on the food sold in the specialised packaging.

So more interesting for the consumer would be a solution that can take food in its current packaging, and deliver it in an automated fashion.

The majority of pets are dogs and cats, and food typically comes in a standard can. The steps an automated feeding device would need to be able to do are as follows:
  • Store a large number of cans, and allow these be used in turn (gravity feed)
  • Open the can
  • Take slices of the can and contents
  • Apply plunger to can slices
  • Eject empty can slices
  • Cover (and potentially refrigerate) can slices
Such a system would allow any subdivision of a can (half, thirds, quarter, etc), based on the dietary requirements of the particular pet. The food would be dispensed into a swappable bowl, allowing it to be cleaned by the user on a regular basis.

Hand belt sander for uneven surfaces

Standard hand belt sanders work well for taking an even amount of material from an even surface. But for taking an even amount of material from a slightly un-even surface, they don't: they even the surface first, which results in a loss of a gentle unevenness that adds character to the surface. This could be solved with a set of additional independently-moving spring-loaded rollers pushing the belt onto the surface. The greater the number of rollers, the more un-even the surface can be. The diagrams below illustrate:

Wednesday, 22 August 2012

Research on the slippery slope

The slippery slope, the thin end of the wedge, the foot in the door, the camel's nose. This argument is used frequently in debate, including rigorous academic debate, despite arguably being an informal logical fallacy. If this argument is to continue to be used (and in all likelihood it will be) it would be useful to have a catalogue of examples to draw from, with mechanisms explained, in order that the use of such an argument can be justified.

Baby farming

I've always been intrigued by the baby farming of Brave New World and other fiction. It seems to me that specialisation and the corresponding efficiency have potential benefits to offer in this realm of human activity (it is one of the very few remaining human activities that hasn't fully succumbed to specialisation). But one lingering concerns I've had is whether a uniformity of service results in a uniformity of the new humans produced, which likely to seriously detriment the functioning of society.

One possibility would be for this activity to be undertaken by numerous private sector (and third sector) entities who would compete with each other to attract business and would adopt innovative (and hence diverse) approaches to child-rearing.

This raises the question of who the service is being provided for. If humans can be conceived and gestated in vitro (not currently feasible, but likely to be with advances in science), and then raised entirely in baby farms, there are no parents (or even adopted parents). So who is the service being provided for? The answer can only be society.

Perhaps the amazing revelation from looking at child rearing in this fashion is that parents throughout the world are effectively providing a free service to society by creating and nurturing the next generation.

What I want from a beat library

Have you ever found yourself listening to a song and wondering how the beat is constructed? Well, depending on what kind of music it is, you may or may not be well served by what's currently available on the web.

So what would a perfect beat library look like: perhaps most importantly, it needs to be complete (every song ever is a tall order, but why not aim high); it needs to have drum score and drum tab for all the beats; it needs to have user-playable files in midi format and proprietary formats such as fruityloops (this allows users to listen to the beat, but also alter it, which helps in understanding what makes it sound the way it does).

This data will also allow interesting comparisons between beats. For example, when you look up a beat the library can tell you "This is a characteristic

Random paper

Hypothesis: the paper we write on, in particular the line spacing and other pre-printed patterns, affects how we're thinking at the time of writing. A diversity of paper to write would stimulate diverse (creative) thinking.

Idea: random paper - it should be possible, even with a standard dots per inch printer, be able to print paper such that no sheet is ever the same. The pre-printed patterns include: line spacing; line thickness; line width; dots; polygons (regular and irregular); non-horizontal lines; other familiar shapes (e.g. the shapes included in common drawing tools); colours; etc.

Testing the hypothesis: how to measure creativity? One method that has been used before asks a respondent to come up with uses for an object - the more uses the more creative.

Tuesday, 21 August 2012

Smartphone lecture mic app

For interactive lectures microphones are often passed around the room or taken from person to person by a teach assistant (or equivalent). This often introduces delay into the process, making the interaction less smooth (and also wasting valuable education time). At the same time, almost everyone in a lecture theatre is probably carrying a microphone on them - on their smartphone. All we need is an app that allows us to connect.

How would such a system work? The lecturer could provide a QR code at the beginning of the lecture allowing all participants to log in to that particular lecture. When the participant wants to ask a question, they press a button on the phone and a notification is sent to the lecturer's computer. The lecturer can then choose to take the question, which makes that participant's mic live (connected to the PA system of the lecture theatre via wifi). Likewise, when the lecturer asks a question, participants can press a button on their phone to respond - will the first respondent's mic becoming live. Additionally, the lecturer can pick a student at random from the set logged in if no-one is forthcoming in answering the questions.

Sunday, 19 August 2012

Support for athletes

Dennis Cotterell is rumoured to have earned half a million dollars coaching the Chinese swimmer Sun Yang to Olympics gold. Few would doubt the value that a good coach brings in ensuring athletes achieve their maximum potential. But why don't we have equivalent support teams (coaches, physiotherapists, nutritionists, etc) behind our top intellectual athletes?

  • Are they less important? No, arguably more so: our top brains are working on things like curing disease, fixing the global economy, and making new products and services that improve quality of life for billions of people
  • Are they less in need of coaching? Unlikely, although not backed up by anything other than personal experience, I would say that it's easier for someone else to get the most out of you than for you to get the most out of yourself
One explanation is that sport is inherently more measurable, in two key ways. Firstly, the performance of athletes is easy to measure (e.g. time taken to do X, distance thrown). Additionally, the cultural importance of the sport is easy to measure (is it an Olympic sport, how many TV viewers does it get, etc). By contrast, intellectual performance is much more subjective.

Another possible explanation is that for intellectual output, the team size can be quite large, whereas for sporting events the number of individuals is in the low tens.

Regardless of these differences, it is a probable and desirable outcome that society expands that coaching skills it has developed into other spheres, particularly the academic/intellectual.

[N.B. leaders (e.g. CEOs, politicians) and actors also have support teams comparable to athletes]

Sunday, 12 August 2012

Learning to get back up

Learning to get back up is the most important thing we can learn. How can we teach it?

Knowing what you know

As important as knowing something is knowing what you know, such that you can compare what you know to what there is to know, and identify areas to focus on.

Setting out what there is to know (i.e. a complete syllabus) is relatively easy, and should be standardised per subject area to enable individuals to map what they know to the what there is to know in a standardised (and hence comparable) fashion.

Syllabuses are typically hierarchical datasets, with subtopics sitting within topics, and subsubtopics sitting in subtopics, etc. However, cross-linking between topics, particularly for dependencies (i.e. prerequisities) is essential.