Roberto Manduchi joined the computer engineering faculty at UC Santa Cruz in 2001 eager to apply his expertise in computer vision to socially relevant problems that could immediately benefit from technology. Previously, he had worked at NASA’s Jet Propulsion Laboratory, focusing on sensors that helped robotic vehicles navigate rough terrain.
At UCSC, Manduchi decided to apply his technological expertise to engineering things that improve the daily lives of visually impaired people. He soon realized that reaching that goal requires more than engineering knowledge; it also involves learning about the lives and abilities of another community. “It’s been a long process of reeducating myself,” Manduchi said.
Now he works through teaching and research to spread his broadened perspective to undergraduate and graduate students.
Manduchi’s first federally-funded assistive technology research produced a "virtual cane" that identified obstacles using lasers as an alternative to the long white cane used by many visually impaired people. Now he considers the project misdirected. “The technology was substituting for something that already worked well,” Manduchi said.
He intended to create a product that helped people, but he had not talked to potential users to find out what devices would be most useful. He has since recognized a pattern that he calls a vicious loop: Academic engineering researchers want to use their algorithms or hardware to help people. The researchers apply for grants, publish papers, and perhaps share their work through the media. However, the devices are not adopted over time, and the researchers return to their labs to develop another product.
“That leads to frustration and resentment from both researchers and the intended users,” Manduchi said. “Researchers wonder why people don’t use their systems. The user sees another useless gadget, a solution in search of a problem.”
Getting to know the community
Manduchi considers his cane project an important experience. It challenged him to learn how the blind community lives and what problems they face each day. By making friends with people who are blind and with those who work with the blind, he’s learned that the blind community encompasses people with varying abilities to perceive light and is as diverse as the general population in terms of age and mobility. By getting to know the community, he has discovered practical considerations that apply to any project, such as the need for affordability, long battery life, and pleasing aesthetics that do not draw attention to the user. Manduchi now continues to gather user feedback during the research process as he develops any new application.
Some of his current projects include helping blind people navigate using cameras in smartphones and helping them aim the camera so they can photograph a page of text, which enables them to use a text-to-speech algorithm that reads the text aloud.
Manduchi takes care to state that his projects are research prototypes; they’re not at a stage where they can be used by large groups of people. “I try not to create hopes that aren’t there,” he said.
Though he speaks humbly about his academic research, Manduchi beams with pride when discussing an undergraduate course he started in 2007 to help students learn about disabilities. The course is taught twice a year, and Manduchi’s fall section regularly has around 200 students. Guest speakers talk about their experiences living with disabilities, and the students work on group projects that get them involved with special needs communities.
This year, students volunteered with Benetech, a nonprofit organization in Silicon Valley that supports technology development for social impact. They proofread digitized textbooks that will be placed online in formats accessible to visually impaired students. Benetech CEO Jim Fruchterman praised the project in his blog: "To better understand the achievement of our UC Santa Cruz volunteers, it’s important to take into account that textbooks are longer and more complicated than the average literature book, and therefore require more proofreading time. Creating an accessible version of a single textbook can take us 90 days or longer and costs $600 or more," he wrote. "The effort of Roberto’s students has a powerful multiplier effect, as the textbooks they proofed will be available to our Bookshare members all over the country for years to come."
Through his course, Manduchi hopes students learn to see disabilities as normal. That’s something he feels is especially important at UCSC because it’s uncommon to see people with physical disabilities around the large, hilly campus.
Other computer engineers at UCSC also focus on helping people with special needs. Jacob Rosen designs robotic exoskeletons to help people regain mobility after a stroke. Sri Kurniawan designs video games to help teenagers get more exercise and to provide speech therapy for children with cleft palates. Rosen, Kurniawan, and Manduchi bring unique perspectives and philosophies to their individual research, but they all value collaborations with non-engineers and user feedback during the design process.
Bringing together collaborators
In 2012, Manduchi started the Maximizing Abilities Through Technology, Education and Research (MATTER) Center to unite the three researchers and their collaborators. Other members of the MATTER center include engineers and psychology researchers from UCSC, as well as nurses, neurologists, and rehabilitation experts from UC San Francisco and UC Davis. Center members have gathered for two workshops, one discussing ways technology can help people stick to a rehabilitation plan and another brainstorming future mobile technology that could be useful for visually impaired people. A stroke survivor and a person with muscular dystrophy attended the first workshop, and Blindsight, a Berkeley-based company working to commercialize mobile technology for the blind attended the second.
Though it’s still an evolving project, Kurniawan says the center is important because it shows there’s a unified front of researchers at UCSC working to improve the world for people with special needs.
Manduchi hopes his engineering graduate students will benefit most from the connections with medical and human-focused researchers in the center.
Developing those collaborations is the only way to develop a product that will have an impact on people’s lives, Rosen said. Collaborations between engineers and doctors, for example, require each group to learn the other’s language. Manduchi hopes the center will help engineering students learn to work effectively with experts in other fields and users of the technology they’re developing.
By broadening the education of the next generation of engineers who will be developing products for people with special needs, Manduchi works to reroute the “vicious loop” that snared him during early projects. He’s changing the future for people who use his products, as well as for researchers developing such products.