Summary: Evolution Robotics provides a variety of technologies and products that support the development of, or can be included in, practical, low cost products that will find wide use in everyday life. What CEMA, Idealab, Vodafone Ventures and Quercus Trust funded was not a robotics company with solid technology, capable management and a proven business plan per se, but a company offering a high level of deep IP, and supplying low cost, practical technology that can be applied to a wide variety of robotics and non-robotics products, largely in the consumer space.
On Nov. 10, 2008, Evolution Robotics Inc. announced that it had raised $13.88 million in a series B round of venture funding. The round was led by CMEA Ventures. Other participants in the funding were Idealab, Vodafone Ventures and Quercus Trust. It was also announced that Faysal Sohail, Managing Partner of CMEA Ventures, would join the company’s Board of Directors. Sohail is a technology industry veteran who has served in executive management positions in a number of high-profile, high-tech firms. He also sits on the board of Applied Wave Research, Catalytic, Pyxis Technology, Inovys Corporation, Symwave, and Telsima.
Finding Its Market
Evolution began as an Idealab startup. In 2002, the company released its first product, Evolution Robotics Software Platform (ERSP), an open software platform designed to be a ‘standard’ for the personal robotics industry. A few months later, Evolution launched the ER1, a ‘personal robot system’ that was designed to be the first in a line of autonomous personal robots. The ER1 system, which required a Windows laptop perched on top of a mobile platform, was largely just that, a mobile Windows PC. ER2 followed, but while the product did catch on with hobbyists, it did not succeed in the larger market.
The ER1 and ER2 products should serve as cautionary example for robotics developers intent on bringing to market, high-end, high-cost (relatively), robotics products targeted to the consumer market that offer little more than mobility and a webcam-enabled PC. It is common for smart toy producers (smart toys are a legitimate and growing market) to add inexpensive consumer electronics functionality, such as MP3 players and sensors, to low-cost, mobile platforms. However, the market for pricey (again, relatively pricey), mobile consumer robotics products that offer little more than their smart toy counterparts (no matter how gloriously designed and well marketed), limits sales to extreme early adopters and robotics hobbyists.
Evolution Robotics has continually updated the well received ERSP product line, a Linux (Debian Sarge, Fedora Core and Red Hat) and Windows based software development kit (SDK) designed for developers of robotics and robotic technology that includes navigation and vision support, as well as a graphical development environment. The vision component of ERSP is provided by its Visual Pattern Recognition (ViPR ) software, while navigation, mapping and localization support is provided through Evolution’s vSLAM technology.
The ERSP product also includes a software architecture for mobile robots entitled the Evolution Robotics Software Architecture (ERSA). ERSA provides application programming interfaces (APIs) for integrating robot hardware with both packaged and custom software modules. ERSA also supports a number of abstraction layers for controlling the robot through both real-time input and event-based behaviors.
Products for the Consumer Market
The ERSP product line has been well received, particularly among academic groups (75% academic discounts are available). Reviewers cite the product’s ability to produce modular, loosely couple code that reduces the complexity of the development and testing process (also rapid prototyping), supports software reuse and hardware independence, as well as increase the portability of the code. Reviewers also note the ability of ERSP to produce lightweight systems (fast with minimized memory requirements) through the use of a variety of managers that load and unload code modules during run-time. In most industries, but especially in the consumer electronics industry where minor variations in component costs differentiate market winners from losers, ‘lightweight’ equates to lower cost.
According to Evolution Robotics CEO, Paolo Pirjanian, the goal of Evolution Robotics is to develop robotics hardware and software that can be embedded into everyday products to make them smarter. The ‘everyday products’ in the preceding sentence largely refers to, but is not limited to, consumer electronics products. That is, robotics will drive consumer electronics. While the ERSP product is quite capable of developing complex, high-end robotics systems, it is clear that the ERSP product is extremely well suited to serve those developing consumer robotics products (lightweight, cross platform, low cost, etc.).
Evolution’s product line is rounded out by its NorthStar technology, a hardware and software localization product that allows mobile devices to navigate more intelligently in indoor environments. Other products and grad student projects are available that can perform the same function. The NorthStar technology differs from the other technologies and efforts by the important fact that it can support indoor navigation at a very low price point… a consumer robotics (or consumer electronics, if you will) price point.
The other key differentiator of Evolution’s NorthStar technology is that it is eminently practical. NorthStar’s ‘Indoor GPS’ functionality is made possible by a tiny piece of hardware mounted on a wall that projects two infrared light spots with different signatures. Detectors on NorthStar-enabled mobile robotic systems, for example a smart toy or robotic vacuum cleaner, can triangulate using the spots to localize itself in a room (or multiple rooms based on how it is configured).
Using NorthStar technology, devices such as robotic vacuum cleaners can locate charging stations and provide better floor coverage in a shorter amount of time. Toys and other systems can run semi-autonomously or along preprogrammed routes depending on the configuration of the NorthStar system.
Partnering With OEMs
In addition to expanding the ERSP and NorthStar product lines,Evolution’s business model also includes partnering with original equipment manufacturers (OEMs) to integrate its technologies into other products. For example, in 2003 Evolution inked a licensing deal with the Entertainment Robot Company (ERC), a division of Sony Corporation (NYSE:SNE), to add advanced object recognition capabilities to their Aibo robotic dog. Similar deals with Bandai (ERSP), Yujin Robotics (ViPR, NorthStar), WowWee Robotics (NorthStar) and others followed.These partnerships have met with various levels of success depending on the success or failure of the OEM product itself. While Aibo was hailed by the public, it was famously dropped in 2006 during a cost-cutting and reorganization effort after selling only 150,000 units. Yujin’s iRobiQ personal robot is sold largely in Korea, and at approximately $5K, it is too expensive to be sold in large numbers. WowWee’s Rovio product, however, appears to be a hit.
Evolution has multiple licensing and partnership deals with toy maker Bandai, as well as with Bandai Networks Co. Ltd, a Japanese mobile content provider, to include the ER Search visual search engine in KDDI Corporation’s au line of camera phones. Evolution has also entered licensing agreements (and sales) with research centers and government groups including the Korean Institute of Industrial Technology (ViPR) and Taiwan’s Industrial Technology Research Institute (ERSP).
During a meeting with Robotics Business Review on October 23, 2008 (prior the formal announcement of the funding), Faysal Sohail, Managing Director of CMEA Ventures, was queried as to what CMEA looks for in potential robotics investments. The answer was direct and unequivocal—he was evaluating companies that offered deep, patented, IP (intellectual property) in technologies that can be applied across multiple products and product classes. Evolution Robotics had that in spades, plus a working business model. What the company needed to take it to the next level was an influx of cash.
It is commonly held that the personal, service and industrial robotics sectors are driven by a confluence of factors, the primary drivers being the increased capabilities of key enabling technology—batteries, processing, memory, networking, etc.,—along with a concomitant drop in prices for those same technologies. It is interesting to note that Sohail believes that the opposite is largely true… robotics is driving innovation in computing, sensors, battery technology, wireless communication and many other types of enabling technology. This same insight was offered by Rod Brooks, Director of the MIT Computer Science and Artificial Intelligence Laboratory, and one of the founders of iRobot (and now Heartland Robotics) in his keynote session at the recent day and a half conference at MIT entitled Robots: Informing Future Design for Human-Machine Interaction in December 2008.
Robotics and Microchips
The second round of funding of Evolution Robotics by Idealab, Vodafone Ventures and Quercus Trust was good news for companies seeking investment through VC channels, but particularly for robotics firms where most VC funding up until this time has been largely limited to the field healthcare robotics.
In a conversation following the announcement of the Series B investment with Robotics Business Review, Evolution Robotics CEO Paolo Pirjanian made a claim similar to CMEA’s Faysal Sohail and MIT’s Rod Brooks, namely that he believes innovations in robotics technology will drive advances in other technologies and products, particularly those targeting the consumer market. Early indications of this line of thought were given at Pirjanian’s keynote at the RoboDevelopment Conference and Exposition in 2007, where he debunked the universally shared belief that the current robotics industry is a mirror image of the early PC industry (Editor’s note: Robotics Trends, the producer of Robotics Business Review, is also the producer of the RoboDevelopment Conference and Exposition). Pirjanian noted that the robotics industry is much more similar to the consumer electronics industry, in that robotics technology will become embedded into everyday devices in a manner similar to microchips. What CMEA and the others funded was not a robotics company with solid technology, capable management and a proven business plan per se, but a company offering a high level of deep IP, supplying low cost, practical technology that can be applied to a wide variety of robotics and non-robotics products, largely in the consumer space.