USC and Max Planck The Double Life of a Top Robotics Researcher

Robotics researcher Stefan Schaal had made it, a top researcher in a field that promised only to grow. But that apparently wasn't enough for him.

Stefan Schaal

Stefan Schaal

By Rex Dalton


When Stefan Kai Schaal decided to earn more money in the future, he took a leave of absence. A long leave. It took the researcher more than two years to integrate his new German life seamlessly and inconspicuously into his old American life.

Schaal's employer, the University of Southern California (USC) in Los Angeles, was accommodating. It granted the renowned computer scientist the sabbatical in the middle of the semester - a sabbatical he had applied for on the day he was thrown out of his home and his wife filed for divorce after nine years of marriage.

That was six years ago. It was followed by a rose war, in the course of which Schaal, who is now 56, had to disclose his income to the Californian court. These and other documents that have been made available to DER SPIEGEL reveal more than just the financial situation of a German computer scientist. They reveal how the globalization of science has made it vulnerable. And they show how dark corners are created in an already confusing system, loopholes that a clever, creative person can use to his advantage.

A person like Stefan Schaal.

At the time, the Max Planck Society was planning a new institute in Tübingen, where Schaal was to become the director. It was a golden opportunity for Schaal, and he accepted the job offer from the Max Planck Institute for Intelligent Systems - but he also held onto his position at USC. Schaal wanted the glamorous post in Tübingen without having to give up the Santa Monica beach. That in itself is hardly remarkable: It is definitely seen as a plus when scientists work internationally.

But from then on, Schaal focused his energy on pretending that each job - both of them primary occupations - was the focal point of his scientific work. And he was apparently so successful that both institutions failed to notice what Schaal was doing. "It only recently came to our attention that Schaal had a full-time paid job at the MPI," said a USC spokesperson. The university is now examining the situation to see if any measures need to be taken.

A spokeswoman for the Max Planck Society, which is funded by taxpayers, said it "would not be appropriate" for a director to hold a second job that requires 100 percent of the holder's working time. At the same time, however, the society views a professorship in the U.S. as a job unlikely to take up more than 75 percent of the holder's working time - essentially a free pass for those interested in holding two jobs.

Schaal himself calls it a "very clean arrangement" between two institutions and that it is a "typical cooperation with complementary research." Later, he said in a statement: "I make sure I fulfill all my obligations to both institutions," adding that the Max Planck Institute for Intelligent Systems pays for his trips to California. Documents show that USC covered his expenses for stays in Tübingen.

But the documents also show that the professor from Los Angeles could not be cloned quite so easily. There were financial irregularities, half-truths and a creative - and wasteful - use of public funds.

Possible Irregular Payments

In 2012, for example, court records show that Schaal was reimbursed twice for at least two trips from Los Angeles to Germany and back. The Max Planck Institute paid $14,000 for the flights alone while USC - using public funds - paid $3,500. According to the records, there were additional problematic reimbursements of various expenses incurred by the professor, such as for his accommodation in Los Angeles after the divorce.

International Newsletter

Sign up for our newsletter -- and get the very best of SPIEGEL in English sent to your email inbox twice weekly.

A lawyer of Schaal's claims that his client reimbursed USC 10,000 euros that same year for possible irregular payments. USC has said it intends to look into the matter.

Schaal's dual-track career, which earns him up to $350,000 a year, including generous expense allowances, shows just how difficult it has become for academic institutions to control their top-level staff.

During his sabbatical, Schaal led a team to develop the control system for a rescue robot that would be able access places in a disaster that would be too dangerous for firefighters and other responders. The Defense Advanced Research Projects Agency (DARPA), which is part of the U.S. Defense Department, launched a competition called the Robotics Challenge. The aim was for participants to design a robot capable of navigating a challenging obstacle course.

DARPA provided funding to a number of competing teams for the development of such a robot. Working together with Sarcos, then a subsidiary of Raytheon, the USC team was to receive $3 million. The winner of the competition, which was scheduled to take place in June 2015 in Los Angeles, would receive an additional $2 million.

But the USC-Sarcos team did not get beyond the first phase of the project, with DARPA cutting off funding in June 2013. By then, almost $1.8 million had already been allocated to the research group. Gil Pratt, who was in charge of the competition at DARPA, says that the grant cancellation was due to design failures.

When asked about the project recently, Schaal initially referred to it as "very cool." Later, though, when the conversation turned to DARPA's suspension of funding, he called it a "stupid thing."

Maintaining the Double Life

Still, Schaal decided to use the robot developed by his team, which had been christened Athena, for a PR coup. In December 2014, half a year before the competition finale, Athena was brought to Los Angeles in a wheelchair and then, surrounded by television cameras, checked in for a flight to Frankfurt. MPI and USC press releases celebrated Athena as the first robot airline passenger in history.

The robot Athena getting ready to board the plane.
REUTERS

The robot Athena getting ready to board the plane.

It is possible, however, that the robot was mainly used to help maintain Schaal's double life. In summer 2013, Schaal requested another year of paid leave from USC. In the statement he submitted with the application, he wrote that the Athena project "demonstrated running two research groups distributed between Germany and Los Angeles is feasible."

It had only been a few weeks since the DARPA money had been cut off, yet Schaal was still bragging about it to his superiors: The humanoid robot developed by his team with DARPA funding would be available for USC and Max Planck researchers, he indicated.

His letter had the desired effect. His direct superior Gaurav Sukhatme, head of the computer science department, wrote he "strongly supported" Schaal's request for an extension of his sabbatical. Sukhatme wrote that the university, in addition to the $3 million in development funding, might also be able to capture the $2 million in prize money from the DARPA competition.

In faculty leadership correspondence, Schaal was described as "a highly respected scholar" and "one of the world's leading experts in robotics." Even during his sabbatical, he had "successfully and fully met his obligations to USC." None of these memos from July and August 2013 mentions that DARPA funding had been cut off. Did Schaal not tell his superiors that winning the prize money was out of the question? That the subsidies had dried up? Schaal says: "My USC colleagues were aware of it without me needing to specifically point it out."

His sabbatical was extended without any problems.

A short time later, the resourceful computer scientist found himself yet another job: as a hobby accountant. In December 2014, Schaal founded SKS Autonomous Systems Inc. in Reno, Nevada. In his divorce proceedings, he stated that he used the company to try to reduce his taxes. It apparently served as a kind of "merchant bank," said Schaal, to take care of payments that were associated with the MPI in Tübingen. Schaal funneled the participation fee for a workshop organized by a colleague in Chile in December 2015 through his company.

Downsizing

Schaal does not deny these transactions. He claimed that every conference organization has such an account and all proceeds were forwarded on to the organizers. Schaal said they "welcomed this cheap financial solution and knew about it from the very beginning."

But Michael Black, the workshop organizer, says that he knew nothing about it. And Schaal was neither present in Chile nor did he have anything to do with the subject matter addressed at the conference.

But his new partner may have. She organized conferences for the MPI in Tübingen before she went on parental leave in September 2016.

Michael Black says that he was unaware of Schaal's full-time job at USC, even though the two men had been co-directors of the Max Planck Institute for Intelligent Systems when it was started. Black also headed the institute from 2013 to 2015.

The man is a model researcher and one who reflects the spirit of the Max Planck Society. In Tübingen, Black developed software that can produce a 3-D model of a person's body from a photograph, which is useful for the clothing industry as well as for law enforcement agencies.

As a spin-off of his research, Black founded Body Labs. In autumn 2017, Amazon bought the start-up, allegedly for more than $50 million dollars. Amazon is now setting up a research center in Tübingen, Germany and is allegedly planning to hire 100 people.

Stefan Schaal has no such lucrative patents and no comparable spin-offs. He says that's not his goal either.

He may, though, have found an alternative goal: downsizing. On the Monday after DER SPIEGEL originally published its piece about Schaal in early March, the Max Planck Society announced that he had been let go. He had, the institute said, "taken on employment that is inconsistent with MPG rules and regulations." And added that the institute was looking into whether it can require Schaal to repay remuneration that he wrongly obtained.

Article...
Related Topics


Comments
Discuss this issue with other readers!
3 total posts
Show all comments
Page 1
bsantina 03/23/2018
1. Administrators loose their sense in the Artificial Intelligence craze
This is just one example of the political and administrational laymen losing their senses when limitless promises are made about AI technology. These are expensive engineering projects, but they are sold by researchers as magical mystery wonder science. Greed of researchers meets naive beliefs and unrealistic expectations of adminstrators. The contrast between the professional management of this kind of research projects by DARPA, and the silly way in which the MPG lets itself be ripped off by a greedy and obviously confused employee should be noted. Employees of the MPG are public servants and should respect their legal obligations.
Emo Todorov 03/24/2018
2. biased article
Some clarifications: (1) it was widely known in the research community that Stefan Schaal had two positions, and if you look at his publications list, his (distributed) group has produced double the output of a regular research group, so neither USC nor Max Planck can complain about productivity; (2) the DARPA program was highly visible and funding decisions were public; (3) the 350k figure sounds a lot, but is not uncommon for senior Computer Science faculty who do consulting; furthermore, people of this caliber make 3 times more when they move to industry. Compare the irregularities described in the article to the amounts of money he is losing by staying in academia.
ghoerl12 03/29/2018
3.
This is an unusually inflammatory article of Der Spiegel, with lots of statements that appear irrelevant or incorrect, if you know how academic institutions work. What is actually the purpose of the article? Is there any criticism that the international collaboration was anything else but very successful and beneficial for science?
Show all comments
Page 1

© SPIEGEL ONLINE 2018
All Rights Reserved
Reproduction only allowed with the permission of SPIEGELnet GmbH


TOP
Die Homepage wurde aktualisiert. Jetzt aufrufen.
Hinweis nicht mehr anzeigen.