As part of the Indie Manufacturing project we are talking to (aspiring) indie manufacturers, suppliers and support organisations to work out what the challenges are, see how people have overcome them and to work out what we don’t know.
We interviewed Patrick Fenner from Push To Talk early on in our research, back in December 2015, as part of checking our assumptions in the challenges or problems they face. Below are the edited highlights of that interview.
Patrick Fenner Interview
Andy: Let’s jump right in. Can you tell us who are you and what are you making?
Patrick: Hi I’m Patrick Fenner from Deferred Procrastination. We are an open source design and engineering company.
We are working on a few different projects including a Nerf firing range for Maker Faire – actually, right now, I’m working on some LEDs to act as a pixel array for animated sprites. We are also making a sensor unit for a training course to teach people how to create an environmental sensor unit. This allows them to better understand how those sensors work. We show them how the black boxes they put in in the field work and what they’re actually doing inside.
Product wise we are working on Push to Talk. Push to Talk is a device that combats social isolation. It’s a button that sits by the side of a service user’s phone and when they press a button they receive an incoming phone call that connects them to another service user who has also just pressed a button. The users can then talk about whatever they like for as long as they like and when they’re finished they just hang up the phone and they press the button again to get another call.
Andy: That’s lovely – how did that come about?
Patrick: I was invited to an ‘Ideas Day’ around Health and Social Care. During the day we went through a number of possible ideas. There was a pitch competition at the end and we won and got some money to build a proof of concept. We built the proof of concept and took it back to them and they were happy so they gave us money to develop a 40 user trial. Then we took it to Innovate UK and won enough money to be able to do a 400 user trial over 3 months as a way of kickstarting the service and allowing us to work closely with an initial user group and test whether it really does what we expect it to do. If it doesn’t help people there’s no real point but if it does what we expect…
Andy: So who are you working with on Push to Talk?
Patrick: We are working with Doug Hopkins, Jereth Cormack and Francis Fish. Jereth and Francis both worked at the time in DoES Liverpool. Jereth’s moved down to London now but Francis is still around. Doug was at the same ideas day and he has a background in Health and Social Care products so he understands that side of the market. Francis is doing the server side software and Jereth is on the design side.
Adrian: So Doug was on your team and you thought something like “oh right, now we’re going to do this thing and it would be useful to get some extra skills and there are some people I know nearby” – is that what happened?
Patrick: Yes pretty much. Being in DoES and knowing what people do and knowing what we needed really helped. I can’t do server backend stuff so it was a case of asking around as to who was available. Obviously there was pay in there which helped. Later on we used somebody else from DoES to do the videos for us for the pitch to Innovate UK and more recently the TSA conference we presented at.
Patrick: The Telecare Service Association. Not the people you find in airports (laughter) – it’s a tele-healthcare body/service.
Andy: So who is the target market? Who is actually going to buy Push to Talk?
Patrick: Obviously the service user gets the button. That is either provided through a service provider – like a health and social care provider paying for it or it may be out of their personal allowance. We prefer the health and social care provider because they’re effectively acting as the gatekeepers for the service then. They are choosing the people in their care who they feel are best suited to it. In that situation we white label the service to that particular provider.
Andy: Can you talk us through the physical design production process you’ve gone through?
Patrick: The first step was pulling together some wireless devices using Arduinos, buttons and boxes and batteries that I got off the Internet. Once we’d got some buttons that worked together in the same room we then needed buttons we could send out for a trial. We did this using Arduinos and Ethernet shields (or Arduino clones). We had shields with a custom board on top with the buttons and the LED’s facing upwards. They are then mounted onto a laser cut acrylic base and then over the top is a polypropylene cover. The whole thing is enclosed so you have a covered volume that’s springy so you actually press the whole of the top anywhere and it will press the button underneath. It has to be as easy to use as possible. LED’s light up the whole thing so you can see the response really easily. We then have a sticker on top so you know where to actually press.
Adrian: So are your design decisions governed by the tools that you had available in the maker space?
Patrick: Physical prototyping is about what I can create with the lowest initial set-up and overall costs. For custom parts there are always initial set-up costs for any custom pieces or I can cut pieces with a laser cutter with very little set-up time. This lets me create hardware relatively cheaply. I think the acrylic base and the polypropylene top in terms of materials each cost under 30p a piece with the only set-up cost being my time. Obviously there would be manufacturing costs into that and we are not factoring in running costs at this point. Even so its method I can use that still allows me to get out enough. If I need to be able to put out 200 hardware buttons in a relatively short period of time I need a flexible, quick process.
Adrian: So you have one eye on the fact that any additional time in assembling is time that you personally are going to have to spend?
Patrick: Yes. Although that’s where some of my speciality lies as a design engineer in that kind of short run production. It helps me to understand what methods and materials are available to try that will get the nicest looking and the most practical thing out with the least start-up costs and least construction time and balancing those two off against each other. 3D printing would take far too long per unit even though it would have a similar sort of cost.
Adrian: So you prefer to design for tools that you have and processes that you can do yourself rather than go and find somebody else who can do those things for you?
Patrick: Yes. Initially I’d looked at vacuum forming some of it myself, but when there were problems with that I reorganised it to use laser-cutting; rather than look externally. That actually made it slightly cheaper – the vacuum former sheets ended up being more expensive – probably the volumes and size I was looking at. I haven’t done loads of vac-forming, so knew more about the laser cutting side of materials.
Adrian: What were the volumes again?
Patrick: 200 hardware buttons but we’ve got a software button as well.
Andy: Is that practical to do that in house? You’ve done it?
Patrick: I‘ve got the parts ready, I haven’t done the assembly yet but yes that volume is practical to do.
Andy: And what would be the tipping point? What would just be too many for you to do?
Patrick: Well 300 sounds like a lot! (Laughter). I probably wouldn’t go more than about 250 although there’s kind of a graduation because you can take parts of the manufacture out, like making boards. I’m not going to make 200 boards – what I did was pass them off to somebody else. You can do a small run of boards and use a pick and place and get them soldered up ready to go. For instance, we have a shield board – that’s 4 LED’s and a button and a number of other components. They are all surface mounted. We get them made up by Ragworm in Kent. At 200 units it becomes enough that it’s worth having somebody else do it. The board costs on that kind of volume are low enough that overall it was worthwhile doing them that way. I still have to solder all the headers on for each one to fit on the top of the board.
Andy: So how did you find Ragworm? How do you identify suppliers in general?
Patrick: Internet searches are always good. Literally an internet search. Actually I think Ragworm was at a show. I’d seen them at a show and then, looked them up on the internet. I’ve previously had boards from Dirty PCBs which is a China-based company which batches up and farms out boards to some of the really cheap PCB houses in China. They can return almost as quick as UK suppliers and at a lower cost but it’s called Dirty PCB because it’s the cheapest – so you can end up trading quality against cost.
Adrian: So your main criteria are cost, turnaround time and quality? Are there any other considerations you have?
Patrick: It’s nice to be able to talk to a supplier – to be able to get feedback directly. I’ve only ever had boards off the Chinese suppliers whereas with Ragworm I had someone I could go and talk to. They were giving me feedback ‘is that right?’ and I could ask ‘am I going to get what I expect out of the other end?’ as opposed to them just doing whatever you say and they give you the mess at the end.
Andy: Did you consider doing things locally?
Patrick: I didn’t know of any particular places locally so it is who you know about. I hadn’t particularly looked either because I’d used them before. A good supplier relationship, knowing I would get the help I needed, saved me time.
Adrian: So you thought “I know those guys, I’ve used them in the past for some similar things and I know it works and that’s better than spending time trawling the internet and talking to other people trying to find out who I can trust.”
Patrick: Yes and from having worked with industrial suppliers in the past, finding a supplier who will talk to you and then return queries in a sensible amount of time is the big first step. When I first started out getting laser cutting quotes one supplier took 11 weeks to get back to me! That is quite a long time. Half the battle is knowing what you need to ask of the supplier or how to word it in the way they expect so they can process it quickly and easily. That and finding the suppliers who are willing to do the levels that you want. So apart from my engineering priorities sometimes it’s just about interfacing between the client and an industrial partner and translating things between the two.
Andy: Were there any ethical considerations when you selected your materials and suppliers?
Patrick: It wasn’t on my list of things to check. I don’t know, the volume that I’m doing is relatively small so we are less focused on the overall supply chain at that point. It did feel less good ordering stuff from a Chinese board house. I’m aware of some of the labour conditions in China but it’s not something I look at as a large part of the decision making process.
Adrian: I’m curious about your materials choice and things like e-waste and sustainability. Does that play into any of the decisions about your choice of materials and the design?
Patrick: There are practical decisions about what’s available and what materials are going to be available longer term. Generally it is more about ease of access rather than specifically focusing on overall sustainability. Partly it is to do with the small volumes we work with. If you want to make more then you’ve got to find the source that’s going to be available for longer or be supplied from a sustainable source.
Andy: So how many do you think you will be manufacturing in the future and what would be a nice scale to operate at?
Patrick: We are a small team—just 4 people—so we’re not actually best suited to be hardware producers necessarily. My intention is to provide reference designs and make available the hardware but the service specifically will open itself up to alternate hardware and alternate software. I’ve also looked at cell phone integration where broadband isn’t available. I have been looking at dial-up modems though getting the on-board modules in small volumes is much more difficult.
We are service agnostic hence the physical hardware that connects by Ethernet and also a mobile app. If another service provider’s already got a button and hardware that they’d like to use then great – if they’ve already got menu space in a piece of software they’d like to use then great – we can register their button interface and make calls across to the API in the same way as any other button.
Andy: So what are your plans for production?
Patrick: Firstly to see how the trial goes and see whether it gets picked up by a health and social care provider or enough independent providers. If it does then the plan it is to build some hardware bundles in volumes of several hundreds or several thousands. At which point we will be looking more towards external production of items even if to get started it means I’m still putting bits of things together.
Andy: What sort of challenges are you expecting to run into when you start scaling up? What’s the stuff that you know now that’s going to be a challenge?
Patrick: The required investment is going to be the big one. Having to buy a load of hardware and services upfront or on supplier credit which is less likely knowing that there’s a relatively long lead time between having to take on the debt and being able to recover that; so particularly looking towards pre-sales as an opportunity to make that a little easier. I don’t see a problem obtaining the kind of moulded physical items and electronics. There’s already a supply chain for those devices and processes. I’m not expecting to run into too many problems in terms of acquiring the services either. There are certain bits of the process that I don’t know fully yet and I need to find out on the way. Upfront investment is going to be a big challenge.
Adrian: You mentioned that one of the problems you face is finding manufacturing people who are interested in talking to you and taking on the work because you are operating at a small scale. Are you worried this might become even harder as you move up to injection moulding or does it mean that you simply need to crank up the volume a notch to make it worth their while?
Patrick: Part of the challenge is to make sure that the processes you use are appropriate for the volumes you want to produce. The difference between vacuum moulding or injection moulding or, on a smaller scale, resin casting means that you are able to re-design and change parts as the volume changes. Your physical device can then morph and change based on what volume you are actually producing at that time even if parts of it are identical. You can always put the same PCB in but the actual form factor is slightly different because the first ones were resin cast and now we’re doing injection moulding.
The challenge is to make sure that the production method is scalable for the number you need to produce. It’s difficult to make the big jump from making 200 to making 1,000 or making 10,000.
Adrian: Do you think that there are benefits in those smaller jumps – is that why you’d look to make smaller jumps or is that just a de-risking the cost?
Patrick: The bigger jump you have to make the more debt you are taking on and the further you can fall. By trying to make a smaller jump you go more slowly – not kind of slower but steadier – rather than going “I’ll take on all this”, I can see if it works and if it doesn’t I’m going home! I suppose it’s been quite nice so far because it’s been pre-funded and we have been able to not hard sell it – if this service fits the service users need then we can continue with it otherwise we’ve have the opportunity to be able to try that and see if this idea fitted right in the world – or adjust it and change it so it does fit. We haven’t got a whole load of capital invested in one particular idea that we have to sell all over. It feels a bit less predatory, a bit nicer being able to mould and fit the environment rather than force yourself into a position.
I’m kind of expecting there will be a load of things that I don’t know yet that will turn up in some way. I can do this lot because I know enough to do this lot and it would be really nice if it scaled up or if somebody picked it up
Andy: You’ve mentioned that you went for Innovate UK – so you’ve had some sort of traditional support – where do you go for information, where do you go for support if you have any and how do you find out about things like Innovate UK and others?
Patrick: They’re the only kind of external funding pots that I have applied for – largely because I don’t like filling in the forms.
Andy: Any other support?
Patrick: No, I don’t think so.
Andy: And what about the Maker Network?
Patrick: That’s much more kind of ad-hoc. I’ve a running joke that I get my Just-In-Time computer science degree from Adrian and in return he gets a Just-In-Time mechanical engineering degree from me (laughter). That kind of cross collaboration of pitching of ideas teaches you a lot. A certain amount of the client jobs I get are just from people who come in to DoES who need the services that I can provide. Just by being in the space and being able to point to other people around the room and say ‘they do this kind of stuff – go talk to them’ or ‘I’m not the best person for this go and talk to them’ or doing collaborations with people and seeing how they work in particularly the programming side of stuff which is not my original area. My background is in Automotive Engineering originally, stress and structural analysis as a speciality and then moving on towards electronics and programming. I had done some programming on computers but not loads and not in any particular language. It is all about the different people with different experience in different areas who are in the same space or going to the same things I suppose.
There are also the shows and things like MakerFaire and Makefest where you see a particular mixing of people from different areas and different sets of skills. Often people are at those to show off something they’ve done. I went with Ragworm as I’d been to a show that they were at and we got talking. There have been ones where I’ve ordered from suppliers – I think it was Kitronik – I ordered from them and rang them up two days before Makerfaire and said can I pick it up from you at MakerFaire?
Andy: So it’s all very informal and ad-hoc?
Patrick: Yes. I can see it’s a little difficult to map because there isn’t a body that represents everybody.
Andy: What do you think the missing pieces are in the Maker community at the moment specifically around people who are looking to do similar things to you as in go beyond being a Maker – a one off – to being an Indie Maker or a Pro Maker or whatever you want to call it?
Patrick: Access to space, equipment and processes. The way that Hack spaces or Makerspaces give individuals and organisations access to a greater value of service than they can afford on their own means that they can effectively pool their resources. That seems to be really beneficial. I originally came to DoES to use the Laser Cutter. I ended up as a technician with a desk. That’s given me access to the workshop and the equipment and the people and the ability to spend time in that environment. I have learned more of what’s there as a result.
Andy: Is it really important for you to be able to acquire these skills yourself as opposed to just pass it off to somebody else?
Patrick: From my point of view as a Design Engineer it’s important that I understand the process that goes on and a good way of doing that is trying the process so it’s not magical. By ‘magical’ I mean it is not given to somebody else and then something appears and there’s some black box process that I don’t really know about. I was looking at through hole plating the other day because I really don’t understand the process of that yet. How do you get some holes on a board that are through hole plated and some that aren’t if it’s a chemical process – does it get masked off at some point? I need to ask somebody who knows something about it because I couldn’t find anything good enough on line other than it’s to do with copper deposition. Although I’m not going to be doing exactly what manufacturers do, it helps to understand it. I saw that even just doing laser cutting – part of the reason why it took 11 weeks for me to get the quote back was because I didn’t know what file formats were expected, what types of preparation of artwork were expected as design drawings… but after I go and use the laser cutter I understand what I need to do in order to run it, so a lot of that is going to be transferrable in a different scale to a manufacturing job or an industrial partner who just does that.
Adrian: It is one of Andy’s bugbears that he’s known me for years and he knows that I want to just learn everything myself. I think it’s on the same level as you because I want to understand it all. Once I’ve done it a few times I’m quite happy to pay someone but I want to know how to do it once or twice so I know I’m not going to get ripped off, understand the terms, etc.
Patrick: Well, I ran the house finances on double-entry book-keeping so I could learn double-entry book-keeping and now I have an accountant! (laughter) I did my first year’s tax returns by myself and now I have an accountant because now I know why you have an accountant and why they’re worth the money and exactly what you’re paying them for. Sometimes you get that from having a partner who has explained it to you. Part of my services as a Design Engineer is being able to explain why I cost what I do and why it takes so many days to be able to do what needs to be done. In the same way I’m looking to understand what somebody else is doing for me when I pay them. Specialised production by specialised partners is ideal but that doesn’t negate me to understand what the process is and the downside of that is you just want to learn everything and that takes all of your time and then you don’t do anything – you just learn a little bit of this and a little bit of that.
Andy: So I asked you what was missing from the Maker community – what would be useful from the Manufacturing or supply side of things – what would be useful to you personally?
Patrick: An increasing openness to very small customers. I realise it’s not something every manufacturer needs to take on but where manufacturers are able to look to smaller customers that is likely to be a market that’s beneficial for them.
Manufacturers who are able to explain what they do a bit more and be able to provide access into what they do makes it easier for somebody to ask them to do something.
In the case of a circuit board manufacturer it might just be making sure there are design rules that you can download for the software that you’re likely to use so that you can do a check to see if the circuit board is expected to be manufacturable . For a person who hasn’t done it that’s a big hurdle that they’ve got to cross.
You worry are you going to look like an idiot when you send this file? It is little things like should all the files just be zipped together in a single file or should they be separate? You have to learn what each manufacturer wants - which layer outputs from the design software are needed, what size of trace is too thin a trace, what size drill size should I use for components, what size should a via size be? I didn’t know a lot of this and I was lucky to find a set of design rules on manufacturers site. It is just a nice list that I can go through and feel a bit happier that what I am sending isn’t rubbish.
Adrian: Yes, I remember the first board I got done was from an on-line place in China and where you just upload your file. At the time I had no idea what I was supposed to be doing! (laughter).
Adrian: So one thing we’ve not touched on at all is certification stuff. Are you planning on getting certified? Are there any issues you are expecting to hit?
Patrick: The board we’ve got at the minute meets the EU requirements for low voltage electronics. Whilst we’re not going through a certification house what we have to do is to show that it does match the requirements. A certification house isn’t specifically required at this point. We use them to take on the liability of whether it’s the right specification so effectively we are self-certifying at the moment.
The boards as they are don’t have CE marks on them but that is because they are prototypes so we can take them back at the end. There’s no requirement for retail so it’s not the retailer taking on the risk – so it’s me as the manufacturer holding the risk under my liability insurance.
Adrian: So they’re still prototypes and you’ve gone through the tests so you’re happy that they should be okay and it’s helps to avoid spending a big chunk of cash to go through a Certification House – not exactly to rubber stamp it but hopefully……
Patrick: I’m avoiding spending the cash on a pre-production prototype by holding the liability – which is a risk.
Andy: So what’s next?
Patrick: I will be flat out making the stuff for the buttons and sending them out for the trial. After that it will be client work generally, whatever the clients ask me to do – from idea to prototyping and small scale production.
Andy: That’s been really good thank you – really spot on.