The Process: How I Made the Beryl Laserlight
Share this post
My life as a hardware entrepreneur started at university, where I designed Beryl’s flagship product, the Laserlight. The Laserlight is a front-facing white LED bike light that projects the symbol of a bike onto the road ahead of the cyclist using a green laser. It helps cyclists be more visible in drivers’ blind spots, preventing cars from turning across their path, which is the number one cause of accident.
The product was born on Kickstarter and is the first in a range of consumer products for cyclists that we now sell around the world. Our proprietary laser technology has been built into all of the shared Santander Cycles in London, and can be found in a growing number of Citi Bikes in New York. More recently we co-designed the new Santander Cycles for London, for which we provided lights, laser, GPS, Bluetooth and sensors.
The original Laserlight began as my university project — I’d dropped out of reading physics at Oxford University to study product design at the University of Brighton. The summer before my final year, having never been on a bike, I decided to cycle the length of the UK for charity. I fell in love with cycling that summer. I returned to university the week after the ride and decided to dedicate my final year to “urban cycling.” Cycling in the countryside had been relaxing and wonderful, but the cities were dangerous and stressful. I became obsessed with identifying and tackling the biggest problem for city cyclists. I didn’t really care what the problem was — getting wet, lost, or having your bike stolen — it just had to be the biggest problem.
Defining a problem
After interviewing many other cyclists, it became clear that personal safety was the number one concern. It was the biggest worry for people who cycle, and the biggest barrier to entry for those who don’t. I spent about six months really getting to grips with the problem before coming up with a solution. I did a deep dive into road safety data and research and carried out human-centred design” research inspired by IDEO’s techniques. This included cycling around Brighton with a GoPro strapped to my helmet and analysing many hours of footage. I worked with a driving psychologist who studied road accidents for a living to better understand the cause of the most common incidents with cyclists.
There was one statistic from all that research that amazed me: 79% of cyclists involved in an accident are traveling straight ahead and another vehicle maneuvers into them. The most common situation is the side sweep or blind spot, where a vehicle just in front of a bike turns across its path. The second most common is when a vehicle in front of a cyclist pulls out of a side junction into its path. I suddenly realised that in both of these situations the threat is in front. A cyclist can see a vehicle ahead of them but that driver can’t see the cyclist.
Then came my “eureka moment.” I was biking around town thinking about this problem when I realised a white van just in front of me couldn’t see me, and I found myself wishing I was just five yards ahead, warning him of my presence. I realized that I could design something to project myself there.
Despite my physics background, I had no practical experience with optics, LEDs, or lasers, so I looked for experts who could help. Fortunately, I met Professor Wang Nang Wang, a leading LED and optical expert, and a professor at Bath University, near my hometown. We decided a laser was the best option for projection. While an LED is a diffuse light source, a laser focuses all of its energy in a coherent light beam to an extremely bright, far-away point. A projected image from a laser can be far brighter than one from an LED drawing the same amount of power. Together, we built the first prototype, lovingly named “fugly” because — even for a prototype — it was f*ing ugly. It was a heavy monster of a device. I somehow managed to attach it to the handlebars of my bike, and it served as a proof-of-concept prototype for my final year exhibition at university. Biking around with it proved the concept’s functionality, but the aesthetics and performance had a long way to go.
I had a difficult time turning that prototype into a product that was beautiful and elegant to use because I had no real design for manufacture experience, and was still trying to design it from the inside out. I started working with a designer named Matt White, and together we worked through dozens of prototypes and design iterations. Matt was able to turn a really ugly prototype into a beautiful product that we could envision in retail.
Along the way, these iterations and prototypes highlighted specific aspects of the design and addressed possible failure points including:
There was a delicate balance between the visibility of the laser projection and the permissible laser safety levels. We wanted the laser to be as powerful and bright as possible while passing all of the industry eye-safety standards. We created the projection of a bike using a diffractive optical element (DOE) — basically a glass lense that has a pattern, nanometres in height, etched onto its surface. The DOE was one of several optical components that had to be aligned before we could we test the final optical power output. We first tested this in the studio ourselves and then later with professional testing houses.
A large aluminum chassis runs the length of the Laserlight and holds the battery, PCB, and optics in place. The chassis also safely disperses the significant heat generated by the LED and laser. While you can model complex heat dissipation in CAD software, there’s no substitute to actually measuring a real life prototype.
We planned to machine the casing out of aluminum for our first batch of Kickstarter lights. However this is an extremely expensive, labour-intensive process. After some research, we learned about a process called deep drawing, which is how a cigar case is created. You can form pressurized aluminum into a thin shell by heating it up and pressing it onto a mold. Although this achieved the more organic and curved shape we originally wanted, it ended up still being extremely complicated and required post-machining for accuracy. With diamond cut edges and sandblasting, our casings ended up being a thirteen step process. We certainly learned a lesson for the future on streamlining our production. I may be biased, but I do think this lengthy and complicated process is in part what makes the Laserlight so beautiful.
Selecting a Laser
I originally thought that we’d use a DPSS (diode-pumped solid state) laser, the same as you find in a laser pointer. However, I discovered an issue with this technology. A DPSS laser creates green light by halving the wavelength of infrared light using a series of crystals. These crystals need to be carefully aligned, and are therefore extremely sensitive to drops, temperature changes, and vibration — not great for a bike light.
Around the time of our Kickstarter campaign, I learned about a new technology called “direct diodes.” Not only are they more compact than DPSS lasers — which are housed in bulky copper barrels — they are extremely stable and energy efficient. However, as a new technology, they were very expensive — around $50 per diode at the time.
This new technology was developed by a Japanese company for laser projectors in mobile phones with the assumption that this would be a widely adopted technology. They had invested heavily in developing them but were left without an application, necessitating a high price to cover costs.
I made the slightly brave and optimistic decision to use this superior technology to build our first batch of lights and we negotiated a price we could just barely afford. We just had to hope the price would drop quickly for future batches. It wound up taking four years to see a significant change in pricing.
I remember flying to China to oversee final product testing and sign off on our first golden sample (the final sample that a manufacturer sends for approval before mass production). I walked into a big clinical room where all the drop, temperature, humidity, and vibration testing takes place. The first machine I saw looked like a giant glass-front fridge that contained metal bars with several Laserlights perched at different angles like birds on a branch. This is where they were humidity testing the lights to check for water ingress.
I was so excited to see the testing process, I didn’t initially realise that this was also the first time I’d see a finished, manufactured Laserlight. Looking at the final product for the first time, I completely fell to pieces. The Chinese testing facility staff looked at me like “Who is this crazy young British woman blubbering in our lab?”
It was incredible because after so many prototypes, the final thing looked so professional. It could be (and was) a real consumer product, with its warning stickers, barcode, and everything.
Despite these moments of excitement, the testing phase with the lab was tough. Because we had no idea what a formalised testing specification looked like, it wound up being quite an ad hoc process, requiring a few trips to China to complete.
Testing at scale
Our products installed in London and New York’s shared bikes have tough demands put on them. Unlike consumer bike lights, they are left out in all weather conditions and seasons, are bashed about by users, and are subject to vandalism. We simulated this in a warehouse, where an enthusiastic engineer smashed a bike into a docking station many, many times to see how our lights held up.
We also ran a trial of 250 units out on the street for a few months to see how they fared before a full roll-out was approved. Back in the office, we got into the habit of leaving units in the shower or fully submerged in a water tank 24-7 to check for water ingress.
Testing throughout the process
Now we do in-house testing at every phase of development. This is a lot easier because we have a full-time team, testing facilities, and a workshop. We can really get our hands dirty figuring out the feasibility of various product details. Our R&D facility is called Lemur Labs (named after our team’s odd affection for lemurs). Here, our R&D team looks at new technology, concepts, and areas we want to explore in three-week sprints — testing hypotheses and presenting what we’ve learned back to the wider Design Team.
When people ask me what qualities have served me well as a hardware entrepreneur, I often say naïve optimism and dogged determination. At least in the beginning, you never really know what you’re getting into and you just need the courage to run at it. I’ve also learned a few key lessons that have made things easier:
Be close to your manufacturing. Getting on a plane to wherever your product is being made and speaking to the engineers is invaluable. I did this to China five times in the first year of production. I had no idea how extensive and involved the design for manufacturing (DFM) process would be. Signing off on final designs and CAD files is just the beginning of a lengthy final engineering stage. There will be unavoidable and unexpected problems that crop up once a product is actually being fabricated. Getting your team and the manufacturing engineers around the same table to problem solve together can save many weeks of work.
Keep retail price in mind. In the beginning, we allowed premium features to creep into the product, including a really expensive laser, diamond cut casings, a magnetic hall effect sensor, and others. It's important to consider the cost of every component so you can hit a target retail price. Every penny saved without jeopardizing quality means you’re one step closer to a viable business. Not having appropriate price margins to offer to large distributors can really impact your ability to scale up down the line.
Be transparent with your backers. At one point in the middle of our first Kickstarter campaign, I realised that we could make a higher quality product (using the superior laser technology I mentioned earlier), but that it would mean delaying delivery. I didn’t know what to do: carry on as planned, but deliver a sub-standard product, or commit to a redesign — which could take many months — and deliver a better product? At a loss for which route to take, we asked the backers what they prefered. I sent out an update explaining the whole situation and gave them the choice. The backers voted, and 98.9% said that they’d rather wait to receive a better product. I was amazed. This was a key moment for us. I was grateful for their decisiveness and it taught me to always communicate openly with the people who support your work..
The truth about making things. Hardware is hard. We have all the challenges of a software company but also have to figure out manufacturing, tolerances, supply chain, quality control, fulfilment, working capital, and a host of other things. In my experience, turning an idea into reality will cost twice as much and take three times as long as you anticipate. Making things for the first time has a steep learning curve, but it is an incredible and hugely worthwhile journey. Holding that final product for the first time, makes all the hard work worthwhile.
- How Kickstarter Creators Are Coping with the Coronavirus
- Kickstarter y el Festival Internacional de Cine de Guanajuato presentan 12 proyectos cinematográficos dirigidos por estudiantes universitarios en México
- Kickstarter and Guanajuato International Film Festival to Feature 12 Student-Led Film Projects in Mexico
- How to Participate in Signs of Change, Kickstarter’s Upcoming Open Call
- Mexican Game Designer Héctor Pérez Funded Four Games on Kickstarter—Here Are His Tips for International Campaigns