With new technologies and cooperation between companies constantly emerging, autonomous driving is fast approaching reality. Yet, many challenges still exist.
This week on Wednesday, July 1st, TuSimple, an autonomous-trucking startup, announced its three-year plan for a nationwide Autonomous Freight Network (AFN). Initially, the company will work together with big logistic operators such as UPS, Penske Truck Leasing, US Xpress Enterprises Inc., and McLane Inc. to create a mapped shipping network designed for autonomous trucks. The AFN will stretch from Los Angeles to Jacksonville, Florida, and consists of four main components; TuSimple’s autonomous trucks, digitally mapped routes, freight terminals, and a system in which customers can monitor autonomous trucking operations and track their shipments.
Founded in 2015, TuSimple is developing a technology aimed at making long-haul trucking cheaper and more efficient. It believes that soon, autonomous trucks will become increasingly important, as human drivers are bound by daily operating limits. Driverless trucks on the other hand, can run day and night, potentially cutting costs and speeding up deliveries. Currently, the company operates in the US in San Diego and Tucson and has offices in Shanghai and Beijing.
Last year, TuSimple raised USD 215 million in a Series D round of financing, bringing its valuation to USD 1.2 billion. It is now looking to raise an additional USD 250 million to invest in the AFN plan, according to TechCrunch.
With this plan, TuSimple aims to almost double the number of weekly freights runs to 93, up from about 50 currently. While this seems achievable at some point in the future, many obstacles remain if it wants to realize the project within the envisioned three-year period.
On the one hand, there are technological challenges. Previously, the company focused on low-speed autonomous solutions for port logistics in China. However, driving speeds on highways are high and the braking distances long. Consequently, the perception technology for the autonomous trucks needs to be able to operate at a much larger distance. This imposes additional requirements on the sensor technology, as well as the software.
On the other hand, there are challenges on the legal side. Laws and regulations in the US governing autonomous vehicles are still in their infancy, and current rules could pose barriers to technology adoption and operations.
These challenges are not unique to TuSimple. The entire industry has been slow-moving forward. Five years ago, many newspapers and companies predicted that by 2020, self-driving cars would be a common feature of our lives. These predictions turned out to be overly optimistic. The main causes of delays are similar to those faced by TuSimple. To train self-driving cars’ AI, vehicles need to be exposed to large quantities of data, which is hard to come by. Moreover, regulators are hesitant to put self-driving cars on the road for safety reasons.
Nonetheless, despite taking more time and money than initially expected, autonomous vehicle manufacturers are making progress and remain optimistic, as shown by TuSimple’s latest announcement.