HVM G+ New Materials 2017 “was an enormous success”

Tim Minshall

• Productivity Puzzle

• UK has a productivity gap -companies are not as productive as they should be (since the

recession in 2008/2009)

• Technology can act as a driver for increased productivity, but needs to be developed

correctly. Development paths needs to be well understood – you can’t just throw money at

the problem and expect solutions, needs understanding of problems/ecosystem/drivers

• New technology isn’t always complicated (e.g. shipping container driving globalisation)

• UK government developing strategy for “4th Industrial Revolution”

• Looking back at how digital tech has developed, we may not even understand/have defined

the problems that additional computing can solve

• Key to link the digital world and physical world – big data/sensors feed from real world to

digital, 3d printing/additive manufacturing go the other way

• Key for additive manufacturing is to move from niche/novelty to productivity

• Value chain – ideas/technical issues, regulatory roles, problems to be solved, and firms who

are there to do it

• Additive manufacturing changes the concept of manufacturing/distribution. Can localise

production, allow customisation/prototyping/tooling at point of usage

• Need to overcome challenges – people don’t understand how to use tech/why to use

tech/what skills are required

• Needs the correct ecosystem – companies/education/policy/finance to ensure technology

can deliver on potential

Krzysztof Koziol

• Carbon shows different properties according to dimensionality (1d vs 2d vs 3d)

• Diverse range of graphene properties can form “platform” material for vast range of

applications

• Graphene ISO only just defined, but term “graphene” has already been applied to variety of

other materials (graphene/graphene oxide, monolayer/bilayer/few-layer, etc.) – properties

vary

• Reproducibility remains hard, particularly between producers (lack of standardisation)

• Other 2d materials form much larger “family”, semiconductors/insulators/combinations of

properties

• Large funding for development (Graphene flagship = €1bn, 4500 researchers, £120m from

UK)

• Lack of understanding of development process/returns causing political rumblings, but

successes not always emphasised

• Production – top down vs bottom up. Top down allows large scale, but small flakes. Bottom

up – more expensive, but higher quality single films

• Challenges now for graphene – quality/consistency, reproducibility, standardisation,

cost/affordability, increasing volume, quality control, End performance

• Standardisation work ongoing – NPL, IOM, BSI, but ISO recently published

• Some of the disillusionment comes from people using the wrong material/using graphene

for the sake of using graphene without full understanding. Sometimes combinations of

material more powerful than solo. Metals + nanotubes + graphene etc

• Applications are now being realised across many fields – sensors, PV, conductors/wires,

biomedical, composites, automotive, heaters, fire retardance, volume applications –

composites, concrete are following, but will require much larger volume

• Again, consistency/quality control is key

Katarzyna Sokol

• View from the research angle – latest developments of energy storage materials

• 2d materials allow efficient storage without expansion/contraction/cycling degradation/slow

charge and discharge rates

• Graphene/Graphene oxide/reduced graphene oxide – high performance/density possible.

• TMOs – stable performance, many cycles produce little reduction in performance/capacity

• TMDs – very high efficiency, even at high charge/discharge rates

• MXenes – artificially produced materials, very high performance and stability

• Polymers – crystalline polymer form “2d” structures. VERY stable (>7000 cycles and <30%

reduction in capacity)

Day 1, Session 2

Madhuban Kumar

• Productivity dropping, even with people working more.

• UK lags behind in the G7

• Reasons – decision making process too slow, lack of structure results in inefficient work,

need for redoing

• Non-productive activities are creeping into the work day – social media/internet/drinks,

smoking, food breaks. Interruptions take ~8minutes to recover from in refocussing

• Productivity gains from automation/AI will be huge – avoiding unproductive problems,

creating new industries/directions.

• Healthcare, manufacturing

• Displacement of jobs, but new jobs will be created – AI trainers, explainers, sustainers

• Need to ensure frameworks are in place – how to ensure correct accounting, ethics

Sandra Stincic Clarke

• Increased data/capacity means that more and more devices can be connected

• BT focussing on connectivity and data handling (sensors input, applications output)

• To link inputs to data hub, need to move across different systems (not always standardised)

to ensure capability. Data flows to “hub” – power is in the data combination from different

sources, and putting that into a useable and understandable form

• Need to get data in as efficiently/quickly as possible, and make it readily accessible for

outputs. Driver for innovation

• Value chain – sensors à sensor data à communication à Data hub à output developers

àend users

• How to ensure value at each step of the chain

• Data hubs form key step – reduces barriers to entry, makes data accessible on

understandable and standardised terms, combinations of diverse data types for innovative

outputs

• Data has no value if no one can access it, or if it is too complex/fiddly to access

• Demonstrator network – CityVerve in Manchester. How to bring all the data together across

different companies to add value

• >100 data feeds – public vs private, open vs closed, local vs city-wide, different data

handling/storage protocols

• Example – bike sharing scheme. City can find routes/usage/speeds/parking/cyclist specific

information on road quality, signage etc. Identify areas for investment/efficiency.

• Sensors/data is already there (100s of millions of sensors), need to interoperate to unleash

full potential and allow innovation

Mel Loveridge

• Major challenges with batteries to meet requirements for flight/EV/renewable energy

storage

• High reactivity of lithium ions limits higher capacity at present

• Battery evolution primarily in anode/cathode chemistry. Fine for laptop/phone/small

portable device. Not yet ideal for EV etc

• Bio-inspired – macrostructures supported by progressively smaller structures/crystals

• Nanomaterials offer many tailorable and promising properties

• 2d materials – large surface area and large interlayer spacing for intercalation – little volume

disruption means less degradation/cracking

• Research now into hybrid materials – 2d alloys, MoS2-graphene hybrids

• SnSi nanowires on electrodes to enhance current collection

• Coulombic efficiency still not high enough

• MoS2-graphene hybrids – interaction with both lithium and sodium for enhanced efficiency

• Growing MoS2 onto graphene – achieving uniform and stable coverage

• Si-graphene hybrids for current collectors deliver very stable (potentially >1000 cycles)

• Power of graphene/2d materials may be greater as additive than standalone material. Need

to explore both 2d materials and interactions with “traditional” alternatives

Cameron Day

• William Blythe has both dedicated product lines and multi-purpose lines for research

applications

Day 1, Session 3

Alex Kendall

• Robots today – only work in very controlled environments at the moment (e.g. warehouses,

houses, closed roads)

• Machine learning – learning through experience/multiple runs. E.g. alpha go

• Why are these not controlling robots for real-world environment? Limitations – computer

vision, safety aspects

• Computer vision – need to teach computers how to “see”. Computers can “image” but need

to interpret/understand. Building from basic focussing to full interpretation and recognition

• A baby needs ~10,000,000 training examples to learn to “see” (~1 year)

• Computer deep learning can take ~3 days to learn to identify objects

• Models can then interpret aspects such as segmentation (object type), instance

(counting/separation), depth from an image

• Examples – self driving cars, drone that can “object follow” while avoiding obstacles

• Questions on this section: how does it learn? Initial is labelled examples, then moves to

unsupervised learning (no labels, just interpretation). Representative training images are

key. Depth perception: looking at pixel movement across frames

• Ethics and machine learning: how to ensure “optimised” systems don’t perform

unexpectedly

• How do you ensure that a model is accurate, non-biased? Ensure it avoids underlying bias in

training datasets

• Avoid “reward hacking” – machines are trained to gain a “reward”. How do you ensure the

AI doesn’t cause problems to generate additional reward. Example – a vacuum cleaner

trained to pick up maximum dust might start damaging house to create dust to gain further

reward

• How do you deal with uncertainty – ensure that you understand the “risks” in the AIs

interpretation

Harry Swan

• Thomas Swan works across multiple chemical sectors

• Innovation platform – popular science/news, university collaboration and deep tech,

talking/working with entrepreneurs

• Financial support/advice/legal/IP/regulations key to bridging from idea to realisation

• Failure is also a key part of innovation

• Recipe for graphene – kitchen blender + graphite + fairy liqud

• Process is scalable, no chemical functionlisation, tunable to deliver desired graphene

properties

• Range of applications – conductive inks, composites, sensing, heating panels

• Enough in the (new materials) market for all competitors acting in niches

Question session

• How to choose innovation direction? HS: key to balance market driven with “deep tech”, be

prepared to kill off unproductive properties

• Ethics/control of AI systems? AK: Products developed in simulation, development through

failure to reduce uncertainty. Innovation needed in regulation – can an AI be a legal entity?

Need for ethical oversight

• Applications for “bottom-up” graphene? HS: market for Thomas Swan is in chemicals, not

machines, so not development angle

• How do you overcome the “good enough” vs “perfect” issue? HS: Follow customer feedback,

follow the money. You can make it better, but it gets more expensive. Quicker, cheaper

products often sell much better. Be clear on what the benefits are for further processing.

• Startup affiliation for AK? AK involved in a startup called “Wave” – developing platform for

autonomous vehicle development

• Patent landscape for graphene? HS: focus for Thomas Swan is on process patent. Patents

can be helpful but can also be a burden. Often preferable to develop the application, then

assess IP landscape. Think of patents as a commercial tool, see if you can get just as far with

negotiation/licensing. Does it justify the cost? Target specific markets where you will be

operating, not just scatter-gun approach. How do you defend as well? Contingency/No-winno-

fee approach can help

• General AI vs AI for specific applications? AK: Not sure how the milestones lead to that.

CNS

• Based in Cambridge and globally

• Converting methane to graphene

• Question – why are graphene manufacturers expanding vertically? – Currently trying to

demonstrate potential or their material, but would prefer to retreat to supplying graphene

to others if possible

Tara Button

• BuyMeOnce

• Background in advertising

• Products like “Le Creuset” – products that are durable, that you can pass down to children.

Looked for website that brought together “longest lasting” products. Didn’t exist, so started

BuyMeOnce

• Found products – lifetime socks, pens with enough ink to last a lifetime, unpuncturable

football

• Launched BuyMeOnce, now hires 9 people

• Vision is to become a “kite-mark” of longevity

• Shift for Tara -promoting the products she wanted, not those determined by company briefs

• Royalty based, so can promote early-stage companies “for free” and support their

development

• Reverse of general consumer market – drive for cheaper means “disposable” economy,

cutting corners means products are less robust

• Traps consumers into spending more (cheaper product lasts shorter, needs to be bought

again)

• Customers can’t see longevity when choosing product – enter BuyMeOnce “stamp of

approval”

• CF energy efficiency label, but for “lifetime”

• The lightbulb conspiracy – lightbulbs deemed to last too long, companies agreed to reduce.

Ethos today is built around “as long as required for consumers to not complain”

• Environmental gains from longevity huge compared to what can be achieved through

optimising manufacturing

• Question – how do we move to a software model – upgrades/maintenance rather than

replacement. Perhaps “leasehold” for household appliances?

• Question – role for standards industry in promoting longevity? – yes, almost always more

efficient to replace modules than whole system. France moving to fine companies who have

poor sustainability

• Comment from BSI – they are also looking into similar, but need input from industry

BSI

• UK national standards body

• Graphene standardisation from BSI: ISO Standard defined (standard on graphene

terminology), published 2017

• Barrier to commercialisation for graphene is standardisation/QC/metrics

• Bringing together experts from government/industry/academia (And internationally) to

understand what needs to be standardised

• Next steps are to understand other areas needing standardisation

• Question: where next? – working group trying to understand. Key areas – Graphene

information (what needs to be provided for sale – method, chemicals, functionalisation etc.),

Measurement standards (thickness, conductivity)

• Question – how long? Need to agree nationally and internationally. Need to ensure that they

are published at the right time (not so early that the tech isn’t defined, not so late that the

industry has moved on). Typically 3 years time period

• Not allowing Brexit to impact for now on international collaboration.

InnovateUK _ Kalyan Sarma

• Focus divided into sectors to streamline growth

• Funding for centres (CPI and equivalent)

• Funding limited, so move to more open calls

• Special Interest group – networking, KTN input, Faraday Challenge fund

• UK Industrial fund ~ £25bn over 4 years

• Ensuring UK technology base for independent UK world

Nick Coutts

• Important to understand where your barriers are to know how to respond to them

• Building value by reducing risk/improving confidence

15th HVM & 4th Graphene New Materials Event – Day 2 – Business Conference Summit

Justin Hayward

• 4th industrial revolution or evolution with acceleration?

• Evocative Telford Aqueduct – ambition, pleasure facilitated – many further great feats of

imagination

• CIR has run 50 full technology conference days – 50 not out – sabbatical

• Nice foresight in 2008 around events on circular economy and quantum computing

• Graphene events since 2013

• HVM events since 2002

• Merge for unique crossovers

Mike Gregory

• Retrospective over 15 years – renaissance of manufacturing

• Realisation that manufacturing is fundamental to economic development

• 2.7-5.1 million people involved in manufacturing in the UK

• Rate of change accelerationg – 4th industrial revolution?

Lord Broers

• Grand Challenges

• Difficult to reduce to 10, distinguished group reduced it to 14 after 3 meetings.

• Engineering award higher than Nobel and Fields is ignored.

Michael Marshall

• Lived in Cambridge from 1932

• Marshall started 1909 – chauffer driven hire cars

• Went to building/maintaining cars

• Then à aircraft in 1929

• HVM since then (Concorde droop nose, enhanced electronics, advanced composites for aero

• Developing field hospitals, transportable medical equipment

• Maintaining C130 (Hercules)

• Apprenticeships key throughout development- – getting young people involved in

engineering

• Have seen Cambridge develop from limited growth (by design – <100000 population) to form

seed for business growth (science parks around the city)

Andrea Ferari

• How long from lab to factory floor? Diamond Like Carbon – 20-40 years, Transistor – 50

years

• Needs time and finance to develop

• Graphene needs to be allowed similar time to develop/deliver on potential – real material,

not software. Engineering take time

• Graphene currently past peak of expectation, now entering trough of disillusionment

• Need to bridge that gap – Graphene flagship ($1bn)

• Target will be applications that are ONLY possible with graphene (e.g. the laser was not

invented to make DVDs, but DVDs only possible with lasers)

• Methods of production developing – CVD now possible on foil rolls (BM Spider), Liquid Phase

exfoliation (dispersions)

• Inks allow capacative sensing, strain gauges, antennae, wearable sensors, 3d printing

• Composites – graphene enhanced. Extrusion possible – helmets, tennis rackets, bike tyres

• Graphene for heat pipes on satellites

James Baker

• Facing the challenge -how do you get graphene “out there”

• Academia forms foundation, but scale-up processes/standards/nomenclature/management

will need to be developed

• TRLs like ‘snakes and ladders’

• Challenges of TRL raising – big steps along the way, need to make sure all levels are covered

• Concept à product

• Need to shift from technology push to market pull (get demand from industry rather than

using graphene for the sake of it)

Martin Agnew

• 10 years development cycle for civilian aircraft, but early opportunities for drones/secondary

structures/space/payloads

• Potential for advanced materials in aircraft – strengthening, dealing with vibrations/stress,

RF/EMI shielding, Optoelectronics, Displays

• Advanced materials enable unique applications. Solar planes for indefinite flight (<50kg, wit

~2kg metal in the aircraft)

• Satellite development – withstanding launch, and space environment

(radiation/temperature)

• Despite long development cycle for whole products, opportunities to engage with us in the

short and medium term 1-2 yrs and 3-5 yrs when it comes to replacement or refit or

restructuring of aspects of larger products

Peter Hansen

• Haydale produce graphene, but are now targeting applications directly

• GBP£ 3.5mn income and a GBP£ 5.4mn order book – listed on AIM as plc

• Coatings and Conductive tracks for sensors/heaters

• Composites for high strength polymers

• Low level addition of graphene to polymers has dramatic effect (65% increase in tensile

modulus), conductive polymers for 3d printing/extrusion

• Addiing graphene to prepregs for strength enhancement

• Conductive prepgreg comppsites for applications such as lightning protection

• Need to balance electrical and mechanical performance, avoid making the resins too viscous

for usage

• Graphene enhanced adhesive – strength vs conductivity

• Key challenge now (as per James Baker) – get the pull rather than the push. Real engineering

benefits needed to build sustainable market

• Need end users to be prepared to pay for development (developers are not charities)

Nigel Bond

• Full digital printing allows variable labelling of labels/packaging customisation (even “batch

of 1”)

• Example – packaging best before dates, production information, serial number requires

individual tailoring

• Baked bean cans, Daily Mirror lucky bingo – used to be individually numbered

• Now printing full colour (since 2010)

• USP is global service network

• Standardised service level worldwide

• 90% market share for tobacco industry due to anti-counterfeit capability

• Inkjet manufacturer/technology company àglobal coding/marketing/marking solution

provider à enabler of digital print worldwide

• Full colour digital printing allows standard labels and variable information to be printed in

single line – value added

• Very strong core competency, not over diversified

• Stuck to knitting for 40 years

• Very high quality service and support globally

• Business located in 120 countries, but products in every country in the world including North

Korea – where it is also serviced!

• Digital advantage – reduced lead time, can reduce waste in packaging (by printing on

demand) – small batch size if required

• Domino cloud allows real-time monitoring/maintenance/control – additional consumables,

failing parts, line efficiency etc

• “Quick Design” allows automation of product information printing (e.g allergens) without

human error. Reduced product recalls

• Bringing Industry 4.0 with traditional printing gives symbiotic and complementary benefits

for users and for Domino themselves

Ramon Borrell

• Xaar initially licensed technology Seiko II, ToshibaTec e.g.s

• Public listing in 1997 to support shift to in-house production

• Focussing on digital printing, 3d printing

• “Industrial inkjet” – anything that is not commercial printing. E.g. additive manufacturing

• Shift to digital printing from analogue printing (e.g. ceramics industry)

• Enter “non-traditional” printing – onto glass/ceramics/textiles/directly onto curved

substrates (shape) often packaging e.g. plastic bottles

• Functional printing growing

• Décor – e.g. ceramics. Not only printing decoration, but glazes (define gloss/texture/grip)],

textured wallpapers

• Printing “direct to shape” allows labels to be removed from process

New technology development in Si MEMS for nozzle/print head development

• High control of piezoelectric material, print control

• Excellent control/reproducibility of nozzle shape/size

• Actuator – 37 deposition steps

• Inkjet offers unique applications

• 3d printing

• 450mm wafer patterning via nanoimprint

• 3d printing – targeting High speed sintering – powder bed with initiator jetted, sinters

powder. 10x faster than laser sintering

Felice Torrisi

• Research firm prediction for wearable technology is for rapid growth to come (from 2019-)

• Requirements for wearable – stretchable (30%) low-power, breathable, washable,

biocompatible (longer term)

• Current wearable devices -typically rigid electronics integrated with strap/support

• For wearable electronics, printing will be key process for manufacturing

• Using graphene thin films, can produce both strain sensors and non-variable interconnects

• Learning lessons from cotton dying to deliver textile printing – want strong adhesion

between ink and textile for washability

• Fully integrated printing systems – dielectrics, semicondutcors, conductors for active device

printing – transistor, logic gate, memory

• Planarisation of textile needed for successful device

• Question – cost? – FT: depends on sector. High cost fashion could absorb price increase

• Question – enviromnemtal remnant? FT – binders are environmentally friendly, nonmetallic,

so no issues anticipated

Session 3

Angelica Anton

• Silk Ventures support companies wishing to expand/enter into Chinese market

• Partnership with SASAC (Chinese state ownership) to access market, find clients etc.

• Learning lessons from Chinese state-owned enterprises as they seek global market share

• Chinese frameworks for industry support internal connections. Companies that are prepared

to fit the framework will have advantages in the market

• Chinese companies are looking to expand globally. Investment/growth power house. Patent

numbers, unicorn numbers, exit values all back this up

• Question re IP – comment: make sure you’re protected. China is getting better about

protecting (and understands issue) but not perfect yet

Paul Cain

• Display industry is about 200 million square metres in total area

Flexenable puts plastic transistors on plastic (printing) for LCD manufacturing

• Technology outperforms amorphous silicon

• Low temperature manufacturing

• Leakage current very low – allows reduced signal-noise ratio for sensors

• Flexible displays – advantage is as much curved architecture as bendable displays/devices

• OLCD has advantages over OLED for large area, long life, high brightness

• Using TAC as a substrate (already used for displays applications ~4 layers per normal display

to support polarisers). 1 billion square metres manufactured worldwide

• Flexenable technology comparable in price to glass-based displays, but with significantly

enhanced functionality

• Can retrofit older displays fabs to produce Flexenable devices

• Scaling up production – partnership in China

• A low Temperature process has knock on benefits

• A very important milestone for FlexEnable, but also one for the industry of organic

electronics.

Question – how so flexible? No birefringence because hghly optimised plastics used as susbstrate

Question – dead space between pixels? 2-5 μm

Question – Touch screen? Yes, possible

Question – Fabless as alternative to licensing? May not fit with Flexenable’s current model

Question – Brightness? Equivalent to glass displays, increased brightness does not reduce lifetime

(unlike OLED)

Question – XRAY sensors? Same size as equivalent glass. Use same equipment to produce

Quesiton – bidirectional bending/dual axis bending? Not explored, but exploring at the moment

Keith Strickland

• Plessey background – semiconductor wafer foundry, later Moved towards products

• Limit of GaN is in wafer scale – too expensive and cannot be grown at large wafer scale

• Now being grown on other substrates (including Si wafer)

• Cost play – GaN on Si

• Issue with growing on Si is lattice mismatch. Solution is to add strain engineering layers

between Si and GaN

• Currently at 6” wafer, but targeting 12” and beyond

• Current main substrate is Sapphire, can’t do larger than 6” wafer without huge cost

• Development of VR limited by brightness/power requirement – role for Plessey with LEDs

• Pixel pitch for VR/AR only really possible with monolithic development for LEDs. Pick and

place <10μm not practical

Gavin Farmer

• Nanodiamond for improved strength/thermal management

• Thermally conductive but electrically insulating

• Surface functionalisation allows processing/usage

• Thermal management in polymers – enhancement vs BN or graphite

• Can work with range of polymers and complementary filler particles

• Mechanism – improved linking between existing filler particles

• Question – mechanical effect? Studying with epoxy for improved wear resistance, but still

early stage exploration

Andrew Williamson

• CIC fills gap in Cambridge sector for Series A funding

• Balance sheet investor – allows longer term investment, patient investment

• Primarily staffed by entrepreneur/deep tech experienced people

• Have coinvested with multiple other organisations

• Investments at all levels – materials, devices, end applications

• Also investments at seed, series A

Sabesan Sithamparanathan

• For many tracking, passive will be necessary –low cost, no battery required

• Remote tracking for buildings, warehouse, stores. Point of sale system, anti-theft

• Initially focussed on aerospace/airport but now in retail, healthcare as well

• PervasID aims to make passive tags work like active tags in terms of

range/detection/positioning

• Current issue is deadspots. Instead use multiple antennae to move deadspots, ensures

complete detection

Scott White

• Follows on nicely from Sabesan’s talk

Phil O Donovan

• Good to have sensibly structured stock options for the entire company

• Find USP

• Hermann only invested in companies that had global intentions (big markets)

• Sold to large company Qualcomm