“[Lu and I] actually had a lot of discussions about it and we talked about this long ago,” Hou added. “Even in the early days when he became the CEO he said, ‘In the end it’s your company, I’m just taking care of it for you for a while.’
Hou, whose previous role at TuSimple was chief technology officer, said the transition was timed for the completion of the company’s driver-out program – road testing trucks without a driver behind the wheel.
“After the driver-out program, we realized that our technical management team has been superior and matured enough that I can take my hands off them,” Hou said.
In addition to promoting Hou from CTO to CEO after the driver-out program, TuSimple also promoted three people from senior director to vice president positions. “This is how much confidence that I have with the team,” Hou added.
As CTO, Hou said he “siloed” himself in the technology side of TuSimple, but after the driver-out program was completed, he reevaluated what role would be best for him to contribute.
“Now we are in a situation where the technology cannot be an isolated part because it is not the highest ROI area for me. The highest ROI area is now the business side,” Hou said. “We need to make sure we’re merging everything together to become a product.”
Hou said his new role is a transformation from working on the technical formulations in creating an autonomous-driving truck to sharing his vision with the business side of the company.
Hou also said he is the person best suited to explain to investors how TuSimple will operate in a commercial fashion.
When TuSimple announced the change in leadership, markets reacted negatively, sending TSP shares down 30%. TuSimple debuted a year ago (April 15) on the NASDAQ at approximately $40 per share and saw its shares trade as high as $79 last summer. But the stock has dropped from $35 to as low as $8.41 per share in 2022 – and currently trades at around $11.
Hou described the market reaction to the company’s leadership change as “stronger” than company officials had expected. However, despite foreseeing a negative reaction to the change at the top, Hou said slowing the transition process was “the last thing we wanted.”
“You give me a chance to go back to that timepoint, we’re still going to do the same thing,” he said.
Hou said he has never cared much about following TuSimple’s valuation. He also encourages his team to ignore fluctuations in the company’s stock price.
“Internally, we know we have done nothing wrong,” he said. “There are times we’ve done something great and the stock was not increasing; and there was a period of time where we have done nothing and the stock went rocket high and then dropped back. We realized this is not an indicator how good our work is in the short term. After all, we are doing a long run rather than a sprint.”
In addition to the long run, TuSimple has been faced with some regulatory hurdles.
In February, the company came to an agreement with Committee on Foreign Investments in the United States (CFIUS) that allows TuSimple to operate with limited CFIUS oversight in exchange for agreeing to keep its U.S.-developed core technology out of China.
Hou said TuSimple’s early fundamental technologies and deep learning were developed the similarly for the U.S. and China but now the specific tech needed to complete development for each country is “dramatically different” enough to make remerging the two sides unnecessary.
“Why? Because you don’t want to imbue one system that is guaranteed to be running in both situations,” Hou said. “What does that guarantee give you? Nothing. So the U.S. team focuses on the U.S. market and the China team focuses on the China market.”
The segregation of the U.S. and Chinese technology allows TuSimple a narrow focus in each market in the race to be the first company with an autonomous commercial vehicle, Hou said, while also pointing out another benefit to the oversight of regulators.
“CFIUS has gone through their scrutiny and we came out clean,” he said. “I think this is a very good achievement, good testimony. Who else can get a testimony from the U.S. government saying that they’re clean?”
Even with “clean” testimony from CFIUS, skepticism over the viability of a company with sensitive technology and ties to China persists.
Last month, reports began circulating that TuSimple was planning on splitting up the company by selling off its Chinese assets. Hou chalked up those reports to market speculation.
In response to the speculation that TuSimple was looking to sell off its Chinese portion of the business, the TuSimple board of directors issued a statement clarifying that the company was exploring transactions at “a subsidiary level for its Asia-Pacific-focused businesses” but that the “exploratory process does not currently include any plans for divestiture or a similar transaction for any of its businesses or subsidiaries. .. This process is at a preliminary stage and no assurances can be made regarding the nature of a potential transaction, if any.”
Hou said he considers the CFIUS decision in February as “a gaining factor” for TuSimple because it allows the Chinese business to be a “subsidiary level fundraising mechanism” for the company.
“The U.S. has a dedicated team, China has a dedicated team, but we’re valued as one company, not two,” he said. “But in fact, we are developing two different things that have separate values in there.
“Look at other Chinese companies raising money from the private market – we’re jealous,” he continued. “We believe we’re much stronger by a huge margin compared to these companies who have high valuations, so why can’t we? So this is a way for us to attract capital – totally from that perspective.”
The capital that TuSimple attracts was one of the sticking points brought up by CFIUS, who in addition to making sure core technologies in the U.S. stay in country, also demanded changes to TuSimple’s board of directors.
Two board members with connections to Sun Dream Inc., a part of Chinese technology company Sina Corp., will step down once their terms are up. Sina’s stake in TuSimple is what initially triggered the CFIUS scrutiny.
Hou said the makeup of the board is not as critical as TuSimple’s technology efforts.
“What is the influence of the board to the company that 90% of the people are working on technology?” he said, adding that he’s seen no influence over the technology from either the Chinese or U.S. governments despite the constant “noise” over tensions between the two superpowers.
“Our biggest enemy is time – we need to get their first,” Hou said. “We’re running in the Olympics and they’re talking about national clashes. We care much less about that.”
Race for Commercialization
What TuSimple does care about is focusing on “making autonomy a real business,” Hou said. “Today, if you’re thinking about operating an autonomous vehicle – no matter if it’s a truck or a car – you’re actually losing money.”
To become profitable, Hou said TuSimple will need to conquer safety and efficiency.
“This is something people have underestimated. Even if you remove the driver, you’re not going to make money immediately,” he said. “You still have to make sure your operational side is foolproof. Without achieving efficiency, you’re not going to commercialize because commercialization requires that you’re making money on every unit, and then you scale it up.”
Reaching the efficiency to produce a viable commercial autonomous truck will require support of tier 1 suppliers for braking systems, steering, engine/powertrain and the onboard computers, Hou said.
Currently, the kinds of hardware needed to create a domain controller for an autonomous truck do not exist, he added, because suppliers who could make them would only do so if motivated by high sales volumes. “So now it’s a chicken and egg problem.”
In January, TuSimple announced an expanded partnership with NVIDIA to develop an autonomous driving controller (ADC) system designed for Level 4 autonomous driving because most current systems available are made for only Level 2 assisted driving.
The ADC is an integral part of TuSimple’s autonomous driving system and serves as an autonomous semi-truck’s central computer that processes mission-critical perception, planning and actuation functions.
TuSimple expects the partnership with NVIDIA will expedite the development of a high-performance, automotive-grade, and scalable ADC that will be integrated into future autonomous truck production programs. This strategic vertical integration will give TuSimple more control over the ADC’s capabilities and accelerate its development timeline.
In the race for commercialization of autonomous-driving trucks, Hou said TuSimple is “the only serious player in this domain.” Other companies have deals with OEMs (original equipment manufacturers) that are only development contracts “without further involvement, no commitment, no marriage and no determinant dates of the delivery of an OEM, mass production truck,” he said. “There is only one contract in the world – the one between TuSimple and [Traton Group company] Navistar – that is real, with a real deadline and a timeline for a commitment to a finished product.”
CEO: Xiaodi Hou
Headquarters: San Diego
Business: Maker of autonomous-driving commercial trucks
Stock: TSP (NASDAQ)
Revenue: $6.26 million (2021)
Notable: TuSimple was the first company to successfully operate a driverless truck on open public roads.