SIEMENS

Throughout India, rows of cookies are marching from baking ovens at ever-increasing speeds. Cigarettes are zipping by at a rate of 150 per second, and production lines originally designed to stamp, squeeze or extrude 500 widgets per minute are routinely being upgraded to achieve new throughput records. Whether its grains of rice, bakery products, or engine parts, nearly everything Indians buy or export starts out on a production or processing line. And to an ever-increasing extent, those lines run faster than any human eye can see. Nevertheless, even as processing revs into a high-speed blur, quality levels continue to climb and prices remain steady or actually decline. How? Welcome to the world of the affordable "thinking" camera.
Thanks to faster, cheaper, more powerful processors and ever-improving, locally-produced, tailor-made algorithms, cameras used on an increasing number of Indian production lines to catch product imperfections are becoming affordable and increasingly reliable. Dr. Mukul Saxena, who heads Siemens Corporate Technology (CT) in India, explains: "It is important that such products be built here in India in order to be price-competitive with products from local companies. For example, we are working with Siemens' Industry Automation Division and a local academic institute to develop a low-cost camera application that will look at the size and polish of each grain of rice as it cascades in a processing center to automatically quantify the quality of the product."
Adds Dr. Zubin Varghese, department head for SMART innovations at CT India, "The algorithms that allow camera capabilities to be tailored to a customer's unique needs are also the key to keeping the price of service and upgrades as low as possible. For instance," he explains, "with regard to a major Indian cigarette manufacturer that needed to improve product quality, all that was needed was a minor modification in the camera's firmware, along with a new image processing algorithm. We performed these adaptations, which immediately eliminated the quality problems — and salvaged our relationship with the customer."

Smart Cookies. Camera technology can also help to save energy. For instance, a major Indian biscuit manufacturer recently switched a range of its automation activities to Siemens thanks to a new algorithm from Corporate Technology that "allows a camera to see each biscuit in terms of color and thickness — two key indicators of baking process accuracy," says Sameer Prakash, who is in charge of marketing for Siemens Automation and Drives' Food and Beverage business in India. "Not only have quality and production throughput increased substantially as a result of this, but the camera information is fed back to the baking oven to optimize its operations — the first solution of its kind anywhere. This has resulted in a five percent cut in energy use."
In addition, he points out that "by being able to customize our intelligent camera's capabilities to the customer's needs, we were able to differentiate our offer from those of competitors. This opened up the rest of the customer's business for us, including ovens, sensors, oxygen analyzers, controllers and human-machine interfaces."
Siemens' investments in microprocessor-based SMART camera technologies are paying some surprising dividends in the context of the world economic downturn. "Keeping in mind the new financial realities, one possible outcome is that businesses may not want to invest in new plants," observes J. K. Verma, Vice President for Business Development and Sales at Siemens' Industry Automation & Drive Technologies Division near Mumbai. "But they are likely to invest in technologies that will improve the efficiency of their existing plants. Our innovative smart camera project in the biscuit business is an example of that. We can help our customers gain competitive advantage through improved productivity and also create pull for our other products."

Sharper Medical Images. Another area in which Siemens is on the road to delivering significant technological innovation through image processing is its C-arm interventional fluoroscopic X-ray system, which is produced in Goa, India to meet the extremely tight price demands and basic needs of developing markets. Here, a thorough analysis of ways to improve the C-arm's camera and image processing module has not only led to improving the system's resolution, but also to a dramatic reduction in price. "We expect to soon transition from an OEM-produced platform that costs around $2,000 to a superior device developed by CT India that we will produce locally for around $500," says Siemens India Healthcare Sector Executive Vice President D. Ragavan.
Adds Dr. Vishnu Swaminathan, head of CT India's Embedded Hardware Systems Program, which is developing the C-arm camera, "The new camera is not a cheap copy of a Western model. We redesigned everything from scratch with a view to cutting costs while meeting the specific needs of local doctors."
Drawing on experience in India's extremely-price sensitive market for surveillance cameras, Swaminathan's team came up with a hardware-software co-design. This included outfitting the new camera with the same inexpensive light conversion technology used in most digital cameras, but bolstering it with additional image processing technology.
Meanwhile, to achieve improved resolution, the researchers opted for a digital video interface instead of an analog one. "The result is that the image you capture and the image that appears on the monitor are identical and that there is no loss of resolution. No competing device in India has this," says Swaminathan. The researchers also plan to outfit the image processing module with DICOM (a standard for digital communication in medicine) and other software that will allow radiologists to flip images or invert colors to get improved contrast or perspective. A prototype of the new C-arm device is expected late in 2009.

Nano Vision. SPM India Ltd in Bangalore, a machine manufacturer that specializes in conveyer systems, is developing the assembly line for the Tata Nano's two cylinder, fuel injection engine. With an expected price tag of about $2,500, the Nano, which will hit the market this year, will be the world's least expensive four-wheel private vehicle. Yet the Nano is not only designed to tap a huge new consumer market, but to deliver outstanding quality. With this in mind, SPM has asked Siemens to develop an extremely affordable camera-based engine inspection, quality control and documentation system. "The system is designed to inspect the automated insertion of the split (snap-in) rings used to hold the engine valves in place," explains CT Senior Research Engineer Thirumalai Kumar.
With algorithms designed to catch inaccuracies in ring placement, and the presence of dirt and other anomalies, the camera is crucial to Tata's goal of accelerating conveyor belt speed from 3 to 4 m per minute to 6 to 12 in the near future. Although still in its pilot stage, the camera already offers a number of advantages.
"We chose it because it's programmable, accurate and supports documentation," says SPM co-owner and director Dr. G.D.R. Krishna. He explains that the camera, which uses optical character recognition to read each motor's serial number and then combines it with an image documenting correct placement of the rings, transmits its images via a wireless local area network to a storage system. "This supports the warranty for as long as the customer requires," he adds.
Having proven the viability of its prototype, Siemens Corporate Technology in India is using its detailed knowledge of the customer's needs to develop a lighting system designed to accelerate the camera's processing time from around 300 ms per image to 200. "This will help meet the customer's conveyor belt automation requirements," says Kumar. "We are working out how to accomplish this while bringing the price down even further."