Robot butterflies – a cautionary tale

Blc-Q2YCYAAMDyGIn his wonderful magical realism book One hundred years of solitude the celebrated Colombian author Gabriel Garcia Marquez describes a scene in which a young girl is surrounded by a cloud of butterflies fluttering around her. She is unafraid and entranced. article-2242598-1653D728000005DC-606_964x577Now imagine that the beautiful robot butterflies designed and made by Festo could be programmed to behave like this, fluttering around in such a way as to transport any human they surround to what might be described as a magical place. How marvellous would that be? Robotic-Butterfly-by-Festo-2You might even imagine that a literary young person who was studying D&T via the OCR GCSE specification might conceive of this as a possible solution to OCR’s exemplar contextual challenge of enhancing users’ experiences of public spaces. What a creative response! And by making contact with Festo the student might even be able to collaborate with their engineers in producing a prototype cloud of butterflies for deployment in a public place such as a park. But what of unintended consequences? Illah Nourbakhsh, Professor of Robotics at Carnegie Mellon University, has written a series of very engaging short stories in his book Robot Futures. They are all edifying with regard to the impact beyond intended benefit of robots in our society. In the story Robot Smog robot butterflies have been deployed in society for just this magical realism purpose but … the way the robot butterflies interact with humans is through eye contact. If you look at one or more of them they will flutter around your head making eye contact. And there is no off switch. They are powered via sunlight. When it gets dark they simply fall to the ground. Once the sun comes up they flutter off again seeking eye contact with humans. This has led to a situation where people walking in the park are afraid to look up and have taken to wearing sunglasses to avoid eye contact. I leave you to read about what else happens. So as with all things technological we need to be mindful of unintended consequences and ‘be careful what we wish for’. In my view Illah’s book would make excellent reading for Year 11 and above. I wonder how often we use these sorts of science fiction short stories to engage our students with the possible downsides as well as upsides of technology?

As always comments welcome


And now this – cyborg dragonflies produced by genetic engineering to act as drones – not exactly biomimicry more bio combination!


Build, Use, Damage, Mend and Adapt – an approach to learning through and about drones

A guest post by Ed Charlwood

What follows describes the work I’ve been doing in school that has led to me to set up a new Drones in Schools Google+ community for teachers.

A convergence of influences

As with much curriculum development, serendipity did its job at the outset of this endeavour, bringing together the opportunities offered by (1) the new GCSE and A Level specifications and their broader content requirements, (2) a growing dissatisfaction with a certain high-profile external “design / engineering” competition that really requires very little design and (3) the discovery of a very interesting little kit. Firstly, the long-awaited publication of the new GCSE and A Level specifications really was a wake up call that we could not continue to plough the same RM / Product Design furrow at either qualification level. I felt it important to embrace the specification in its entirety and that meant that at Latymer we would have to teach areas that were less familiar i.e. Systems and Control and Textiles. It also meant that we could fully embrace previously fringe areas that we had been pushing at for a few years but had been confined by old assessment criteria, namely the use of CAD, CAM and the circular economy. Secondly, I have seen our students be equally engaged and frustrated with external engineering competitions, they promised a glimpse into the competitive world of high level engineering but actually offered little real decision making, restrictive and difficult manufacturing processes and actually required a lot of luck and frivolous administration. I won’t name names. Lastly I came across a $99 / £78 kit from Flexbot, offering a 3D printable drone and the promise of an open source kit. A quick PayPal purchase later and I was the proud owner of a Flexbot Quadcopter (4 rotors), cleverly packaged, with a comprehensive and appropriate information booklet and a product that worked pretty much straight out for the box and could fly via an iPhone app. Bingo.

Drones are a great ‘hook’ for learning

Drones are popular in the media, comprehensible to most people and on a steep curve of becoming demonstrably better and cheaper at the same time. Currently they have the elusive “engagement factor” and this provides a ‘hook’ making them intrinsically attractive to students. Such a hook is, in my experience, vital. It is important to note that we are not coding experts, nor are we overly interested in programming. But we are interested in using electronics to do stuff. And it is here that the Flexbot Quadcopter meets our teaching intentions.

Our approach

Under the guidance of my colleague Nick Creak we handed the kit over to our students. They assembled the drone without difficulty. Then they had a play, crashed it and naturally broke it. They took the kit apart and made some key measurements, download CAD files from the Flexbot Wiki (SketchUp) and Thingiverse (.stl) and printed a replacement for the part for the one they broke. They then began to explore the files and started to design their own drone. Initially they did this by pretty much by simplifying and copying the existing design, a useful process in its own right to develop CAD techniques and collaborative skills.

A 3D printed Flexbot part

We then printed their chassis designs and used the slicing software to investigate various manufacturing options:

  • How long would the print take if it was “ultimate” or “low” quality?
  • What would happen if it had a low / medium / dense fill?
  • What were the implications of the design being aligned differently?

On average a “normal quality” high density print would take 2 hours. The booklet provided by Flexbot also has some interesting text comparing the economics of 3D printed manufacturing vs mass production techniques like injection moulding.

Students then could begin to design “iteratively” – a new key concept in the OCR interpretation of the new specifications.

“Iterative design is a design methodology based on a cyclic process of prototyping, testing, analysing, and refining a product or process. Based on the results of testing the most recent iteration of a design, changes and refinements are made.”

We also offered a number of design challenges: design a modular drone, alter your design to use as little filament as possible (make it cheap!) or to print as quickly as possible, design your drone to use a standard component – in our case this was a Lego axle.

Flexbot parts

The Flexbot circuit is robust enough to be shared between students and the batteries, propellers and motors are cheap enough to buy in bulk. If you do not have a 3D printer, jobs can be specified, costed and outsourced to a 3D print hub. The simulator (which is available once you have started the process of uploading parts for hub to print) shows it would cost approximately £6 for a basic chassis made from PLA by Fused Deposition Modelling. Some hubs even offer 25% student discount and most do almost next day delivery.

We additionally posed a number of extensions questions to our students, each eliciting a different design outcome: What is the effect of changing the alignment of the rotors? How big/small can the drone be? How much weight can it pick up?


Design Decisions Pentagon

David Barlex has produced a design decision pentagon to describe the decisions that students might make when they are designing and making. So I was intrigued to use this to explore the decisions that our students were making.

Clearly they weren’t making any big conceptual decisions – the sort of product had already been decided – a quadcopter drone. The technical decisions in terms of how it would work had also been decided – four electric motors linked to flexbot circuit, controlled by the Bingo app. But there were lots of possibilities in the constructional decision-making.

Not 90°!

One student changed the alignment of the motors so that they were no longer at 90o to one another which made the drone faster but harder to control. And I suppose you could argue that this constructional change did in fact change the way the drone worked. A key feature of the pentagon is that the design decisions featured at each of the vertices aren’t independent of one another hence the lines between the vertices.

Interference fit

Another student responded to the modular challenge producing a design with four separate arms held tightly by an interference fit to the central node, taking advantage of the high degree of dimensional accuracy of additive manufacture. This required investigation and was in itself was a valuable learning experience.

Clearly it’s possible to set particular design challenges around constructional decisions e.g. making it more crash worthy.

Aesthetic decisions could also be made. Indeed changing the alignment of the motors could be seen as an aesthetic as well as a constructional decision. Devising light-weight covers that can be 3D printed or perhaps produced from nets that have been laser cut from thin sheet plastic might give the drone different ‘personalities’ and this may be seen as a marketing decision, changing the appearance to have appeal to different users. Marketing decisions can also be made with regard to how the drone gets to market – via a kit in a shop or on line, or via digital files for home or hub manufacture in collaboration with a circuit board/electrical motor supplier, related to this, deciding whether the product is open source or not is also a marketing decision. And just who the drone is for will make a big difference to what it might look like and additional features. And taking a step back how will the design decisions overall be affected by requiring drones to be part of a circular economy?

There is, of course, a “purer” engineering challenge, to design and make racing drones, where there are already a number of competitions with related rules and constraints.

The next area for us to consider is that of the consequences of drone technology, and its close cousin the Unmanned Aerial Vehicle (UAV) many of which have some more sinister applications; bombing, surveillance and smuggling as a counterbalance to the positive aspects; photography, delivery, surveying etc… each is a rich seam for discussion as well as the wider issues of automation, disruptive technologies generally or government regulation and control.

Far from this being a proprietary endeavour I want this to be a collaborative, open source one, so I invite you to join the Drones in Schools Google+ community to share your experiences, ideas and resources or add your comments to this post.

Ed Charlwood headshotEd Charlwood

Head of Design & Director of Digital Learning at Latymer Upper School, London

I am a passionate advocate of Design education who believes in the power of learning through analysis, designing and making. I am an Apple Distinguished Educator (class of 2013), a Google Certified Teacher (class of 2015) and the DATA Outstanding Newcomer to Design and Technology Award winner (2008), a particular focus of my work is to exemplify the notion that innovative and appropriate use of technology can redefine the traditional teacher-learner relationship and transform educational designing and making experiences. My vision is to inspire and empower students to make the things they imagine.

Baddass Biomimicry Part 2 Science fiction becomes science fact!

skeeterHave you ever watched a dragonfly? They can hover almost as if frozen in space wings beating so fast they appear as a blur, land with delicate precision on a waving blade of grass, skim gracefully over a pond and fly off at speeds that defy sight. Surely a target for biomimicry and that of course is what has happened. A helicopter used by sea rescue services based on dragonfly flight would be wonderful. Hmmm, scaling up insects is tricky. The fossil record indicates that the largest flying insects existed some 275 million years ago had wingspans of only around 700 mm (28 inches). So may be a dragonfly based rescue helicopter is conceptually inept. So in this case biomimicry has to stay in scale. In which case if you could mimic a dragonfly or aspects of a dragonfly what would you mimic. Given the aerial dexterity of the dragonfly it’s not surprising that Animal Dynamics, an Oxford University spin off, has developed Skeeter a tiny flapping winged drone specially designed for covert surveillance. Weighing no more than 30g, and designed to cost less and fly for longer than other hand-launched drones, it could, its creators claim, help reshape urban warfare. Biomimicry transforming urban warfare! It’s not difficult to see biomimicry playing out in armaments developments. Should this be discussed in D&T lessons? On the grounds of the subject reflecting activities in the world outside school it is difficult to say ‘No’. But any discussion will move into tricky territory very quickly. A surveillance drone, even a tiny one, can easily provide targeting information and missile flight path data for larger weaponised drones. And without too much difficulty be developed into a lethal weapon in its own right. Some argue that the basic technology itself is has no moral compass. The guidance technology used in missiles can just as easily be used for autonomous farm equipment. Where does this leave the designer? And where does it leave the design & technology teacher? As always comments welcome.

Baddass Biomimicry

antEarly in 2013 when there was considerable debate about the government’s proposed National Curriculum Programme of Study for design & technology. Dick Olver, chairman of BAE Systems, one of the UKs biggest companies, criticised the government’s proposal on the following grounds: The draft proposals for design & technology did “not meet the needs of a technologically literate society. Instead of introducing children to new design techniques, such as biomimicry (how we can emulate nature to solve human problems), we now have a focus on cookery. Instead of developing skills in computer-aided design, we have the introduction of horticulture. Instead of electronics and control, we have an emphasis on basic mechanical maintenance tasks. In short, something has gone very wrong.” The result of such authoritative criticism was a complete revision of the proposed programme of study such that it included the following statement under the teaching of design: To use a variety of approaches, such as biomimicry and user-centred design, to generate creative ideas and avoid stereotypical responses. Although biomimicry was a non-statutory example of a design strategy it was mentioned by name.

The Design and Technology Association ran inset sessions to help teachers understand what was for many a new idea. And many teachers have since taught pupils at both KS3 and KS4 about biomimicry, particularly how designers have used it as a creative product design tool. At its most basic the development of webbed gloves and flippers to aid swimming (biomimicking a frog) and more sophisticated the use of corrugated card for a cycle helmet based on the bone structure in a woodpecker’s skull. And of course it’s possible to view the circular economy as a systems approach based on biomimicry that can be used to move the world away from a destructive linear economy.

Kill Decision CoverUnderlying this appears to be the idea of biomimicry as a benign design tool; one that can only be used for good with few if any harmful consequences. But this view misrepresents nature and the constant struggle between and within species for survival. This was made very apparent to me when I read Kill Decision by Daniel Suarez. It’s a rollicking good read but I won’t go into too much detail as this will spoil the story for those who haven’t yet read this excellent piece of science fiction which borders very much on science fact. A key element of the story is to use biomimicry of weaver ants to develop swarms of lethal quadcopter drones that once unleashed can operate without human intervention and control. Weaver ants are able to communicate with one another by laying down and following pheromone trails which indicate the task to be accomplished be that foraging or territory defense. In the case of territory defense the trail will lead more and more ants to the sites where defense is necessary and even large intruders are soon overcome by the multitudes of smaller weaver ants that converge on the site. The brain power of individual weaver ants is of course very small but the colony achieves highly effective defense by getting large numbers in the right place at the right time to attack and kill the intruders. So imagine using biomimicry to transfer this ability to a swarm of drones, each drone with highly limited AI and equipped with simple but effective weapons.

Noel SharkeyThis led me to ponder the role of design strategies in general. In themselves they might be considered neutral in terms of being intrinsically good or bad but their use will of course depend on the intentions pursued by the designer. So the buck clearly stops with us humans. The case of robots and the intention to use them in warfare has led Noel Sharkey, Emeritus Professor of Artificial Intelligence and Robotics & Public Engagement University of Sheffield, to urge extreme caution and argue for international conventions to govern their development. So as always with design & technology we find ourselves in territory where values are as important if not more so than knowledge, understanding and skills.

Stuff we like about Robotics

Robotics VSI2We’ve just added a page of Stuff we like about Robotics. We want to provide some background to the Robotics materials we have published for teachers as a part of the Disruptive Technologies project and so we’ve gathered together some books, websites and other materials that we have found useful in developing our thinking about robotics as a disruptive technology.

To help with navigating what turned out to be quite a large collection, the resources are grouped into what we hope are useful areas for teachers:

Start here
Teaching robotics
Influences on robot design
The Personal; how robots and humans shape each other
The Social; the impact of robots on society
Building robots

The aim is that stuff that appears on this page will remain relevant over a reasonable time. However there is also a constant stream of robotics news at the moment, much of which is ephemeral but could be of use in your teaching. Clearly, if you have the time and interest, you could pick up a lot of these yourself by visiting sites like the BBC, Wired or MIT’s Technology Review. More easily (for many) you could also follow David on Twitter as he often shares these kinds of items. In addition to all of that, we’ll occasionally publish on this blog a compendium of robotics news items that we think might be intriguing, surprising or otherwise compelling.

If you have recommendations for resources that should appear on this page, or interesting news items that we might blog about, please do let us know.

Commons enquiry in Robotics and Artificial Intelligence

Robotics TGThe House of Commons Science and Technology Committee are holding an enquiry into Robotics and Artificial Intelligence. David and Torben submitted written evidence which says, in a nut shell, ‘teach young people about them at school through D&T lessons that encourage them to consider the consequences of deploying technologies’. Clearly our Disruptive Technologies Project is aimed to help teachers do just that. And as if on cue The Robotics Teaching Guide is now available here, later than expected but we think you’ll find it useful. As always comments much appreciated.

Robotics Teacher Briefing – available earlier than expected!

Robotics TBTo complement the Disruptive Technologies Teachers’ Guide we are producing a Teacher Briefing about each individual disruptive technology. Each briefing is organised as follows:

  • Sections 1 and 2 discuss the disruptive technology in broad terms.
  • Section 3 considers two or three particular examples in some detail.
  • Section 4 discusses how the technology might be disruptive using the McKinsey criteria for disruption.
  • Section 5 considers trends in uptake and impact.
  • Section 6 discusses contentious issues that might arise in relation to the deployment of the disruptive technology.
  • Section 7 discusses briefly the interaction of the technology under consideration with other disruptive technologies.
  • Section 8 lists useful web references that will allow the reader to keep up to date.

The Robotics Teacher Briefing is now available.