The technology pusher
There are probably many products out there that would never have seen the light of day if their inventors had listened to market research; the wheel, for example. Until you see it, you don’t really know you want it. Somebody had to be the ‘technology pusher’, putting their ideas out there and letting the market pass judgment.
Point Grey Research prides itself on being such a pusher. Its roots are in academic research and it started by making stereo cameras for other academic researchers. It turned out that the VGA camera with automatic synchronisation and FireWire interface was actually what a lot of people were looking for, only they just needed one of them.
So, while in the back room the engineers are developing esoteric stereo and spherical cameras, in the front room they are knocking out a range of volume cameras with constantly evolving features. A key feature is the ability to significantly upgrade the camera in the field with an emailed software patch. The huge range of features has made the camera a hit with OEMs, who can effectively get a custom camera for the price of a volume product – as they are happy that it comes with any interface you like, as long as it’s FireWire.
Point Grey Research was founded 10 years ago by a group of research students and staff from the University of British Columbia. They had been doing research on computer vision, rather than industrial machine vision. They had a small grant and a licence to commercialise some of their research results, but other than that they had no capital. In the early days it was run from a front room and they tried to start revenue flowing by taking on consultancy jobs.
The first product was a trinocular analogue stereoscopic camera, which was designed for other academic research engineers involved in mobile robotic vision. Point Grey director Vlad Tucakov said the company was focused on developing technology and then trying to find a market for it, rather than the traditional business school model.
He says: ‘We have always questioned the MBA books. We had zero finance; we just reinvested what we could make through hard work and it’s got us where we are today, which is 50 people in two continents with offices and manufacturing – and we are growing at a very fast pace. We have no interest in going public or selling out. We just want to have a proper business that keeps the shareholders and employees happy. The shareholders and employees are, of course, the same people.’
The key ‘value add’ of the stereoscopic camera was the software to calibrate and synchronise the cameras, which saved researchers a lot of time. This later became a pretty useful technology investment. The academic users turned out to be very helpful, because they were also prepared to help debug the software even though they didn’t have a lot of money to spend.
Tucakov says: ‘Eventually we realised that the analogue cameras that we were able to afford to buy were no good for what we wanted to do, so we decided to start building our own cameras from the CCD sensors. We always saw it as a combination of hardware and software solutions so we took ownership of the hardware, firmware and software. That way, if there was a problem, we would have to fix it. That was a very compelling thing for customers who were buying cameras and getting no software support.
‘We made a fully digital version of the stereoscopic camera with no frame grabber and, when that went out, we started getting calls from people saying “this works really well, but I don’t need these two extra cameras”. There were only a few companies making FireWire cameras at the time, but we found it really good for actually getting things done. What is nice for us is that it’s really hard to implement. You need a lot of engineering resources to come up with a good FireWire camera.
‘Out of the gate our FireWire cameras had a lot of extra features, such as automatic synchronisation. Back then, people wondered why we did it. There wasn’t really any demand for it at the time and nobody was stamping their feet saying we must have it. But then people started coming up with multi-camera applications. What a dream for a camera manufacturer that people now order 12 cameras at a time. We knew the value of it from our stereo vision background, but nobody really asked us for it.
‘Our customers are engineers who need a solution for a problem. Our advertising is geared towards telling engineers about the features that we offer.’
Over the years, Point Grey has tried to come up with a range of products that are aimed at OEMs. Rather than build a different camera for each OEM application, it has incorporated a wide range of features into the standard product line. Most customers may only use a few of those features depending on the application.
Tucakov says: ‘There is a lot of reuse of firmware. We were one of the first companies to offer infield firmware upgrades. We can literally email you a new camera. Our cameras are designed to be fluid and use a lot of FPGAs and microcontrollers, so there is a blurring of what is hardware and what is software. This means you can add features to cameras that are already out there, but you can also reuse these features later on in the standard product range.
‘There is no price penalty for the features that are jammed into there. We did consider selling the software in modules, but we decided to give everything at a reasonable price.’
Point Grey has a range of cameras that can either be board level (Firefly and Dragonfly) or boxes (Flea and Scorpion) with either 1⁄3- or 1⁄2-inch Sony CCD sensors. It recently added the Dragonfly Express that has a 1⁄2-inch Kodak sensor, which can produce VGA output at 300fps. The only interface is FireWire (1394a and b) and Tucakov says that this is likely to be the only interface standard for the foreseeable future.
He says: ‘We are not interested in Camera Link. We are not even considering USB or GigE, because it would just defocus us. In a similar way, we are not going to go into optics or lighting.
‘We went down the road of developing CMOS cameras, but they just didn’t work like a CCD camera. It’s not that they were not good enough for the application, they were just a little bit different. Customers were just uncomfortable out of the gate. We found that, with a range of Sony CCDs, the customers get what they expect. We are going down the CMOS path again with the Firefly MV, but it smells like a CCD and has a global shutter. It sells for 100 bucks. It has a place like everything else.
‘You can’t kiss all the pretty girls. There are manufacturers out there who do every conceivable kind of camera with every conceivable interface and none of them very well. FireWire is so good – and hard to get into.’
It also has a range of stereovision cameras and that is still an important part of Point Grey’s business. Tucakov says they are not actively marketed, because they already know who is likely to be interested in them and they talk to those customers directly. He believes that stereovision will become more significant in years to come.
The third product family is spherical cameras. Again the markets are limited to defence, security and a small number of other applications, like entertainment, and again Point Grey is well known to the potential customers already and vice versa, so there is no need for active marketing. Stereo and spherical imaging comes from the academic research roots of the company and Point Grey is quite happy for this part of the business to fly under the radar.
In the early days Point Grey relied on contract manufacturing, but over the years it has created its own facilities. Tucakov said it helps the company to remain price-competitive with relatively low volumes of products. There is no point in moving production to the Far East, because the products change too quickly. It is more important to have a flexible production line that can respond to customer demands quickly. The investment has been funded entirely from the proceeds of sales.
Promotion of the brand has never been an important part of the company’s marketing. Customers have found the products through internet searches rather than being told about them through a local reseller network. Point Grey has expanded internationally with distributors in the Far East and, about a year ago, it opened an office in Munich to serve the European market. But it sees the overseas presence as being part of the relationship building with customers, rather than a sales channel.
Christian Loeb, who runs the European office, says: ‘Our customers are OEMs and they like to have someone in their time zone. They know that it is a Point Grey employee they are contacting and they will get the right answer straight away. They also want to deal with someone who understands the language and culture that they work in.’
Tucakov says: ‘We like to have control. Distributors tend to push what they make the most money from. Also, we are not in a rush. You hire a distributor because you need to buy yourself the market. A lot of customers think that the manufacturer doesn’t care about them, because they don’t even know the customer exists.’
Tucakov is vague about the company’s plans for the future, simply because with no external shareholders, they feel no need to come up with fancy business plan for the next five years. He says Point Grey intends to carry on doing what seems right at the time. The R&D effort is less blue-sky than it was when the company started, but he still believes it is up to the manufacturers to come up with new ideas that the customer may not know they want, as well as listening to customers about what they say they want today.
He says: ‘You have to wonder sometimes whether what customers ask for is actually what they need. We are the experts on the cameras. Customers often come with the solution they have come up with, rather than the problem. We like to know what the problem actually is, because what they are proposing may not actually be the best possible solution to the problem. Customers often say “you’re not listening to me” and I tell them they are not listening to us either. For us, it’s all about relationships. Of course the customer is always right, but we believe that there has to be a relationship between the customers and us, because when you are in an OEM business you are both trying to win the same customers in the end.’
The individualistic approach is one of the luxuries of having no external shareholders. The company is run by a board of directors, who are the five original founders of the company. There is no CEO and each director has a specific area of responsibility.
Tucakov says: ‘Many customers and even new employees find it different at Point Grey, because there are no investors, so there is no pressure to meet some story that can never be achieved, but “we told it because we needed the money at the time”. This just does not exist here. We try to do what makes sense. We still have targets and we keep an eye on performance.
‘Decisions are made by the five founders and we make them by consensus. We try as much as possible to include other people, like our managers in Europe, the US and the distributors in the Far East. I am responsible for sales and marketing and other guys are responsible for engineering and manufacturing and software and support – we are the three managing partners. We recently hired someone who asked what her title was, and I said “What do you want it to be?”
‘It is very clear how decisions are made; there is no confusion here. There is some confusion on the outside. People know who they report to and how things are going to get done; it’s just that we don’t have titles.
‘Investors want some guy they can fire when things go wrong, because they can’t fire everybody, but we don’t have investors. Some things do take a little longer. To be honest it works because we are the size that we are. As we grow I would imagine that a lot of things will have to be formalised.’