By all appearances, May Mobility was a scrappy success story. The autonomous transportation startup made its debut at Y Combinator’s demo day in 2017, with a team that had been working on driverless tech since the third U.S. Defense Advanced Research Projects Agency (DARPA) Grand Challenge in 2017. Within the span of a few years, May had a roster of paying customers in Michigan, Ohio, and Rhode Island as it raised tens of millions in venture capital from investors including Toyota and BMW.
But inside, it was a different story. May’s engineers struggled to maintain and upgrade the company’s vehicle platform, and spent months looking to install an air conditioning formula in midsummer. Management’s ambition has exceeded May’s ability to deliver, which has disappointed suppliers, some of whom have not received period payments. And none of the company’s industrial routes have come close to full autonomy.
Conversations with former May workers reveal that a startup is suffering from standing out in an industry governed by traditional operators such as Waymo, Uber, Aurora, Cruise and Amazon Zoox. As one source said, May’s goal might not have been malicious: the leaders at the summit were convinced it would succeed. But above all, responsiveness and inexperience have led to mistakes that have spoiled municipal relations.
May Mobility was co-founded through Edwin Olson, who in the past was a principal investigator in Ford’s driverless car program and co-led autonomous driving projects at Toyota Research Institute, Toyota’s mobility and robotics division. An associate professor of computer science, Olson took over as CEO in May after taking an absence permit from the University of Michigan, joining GM Ventures veteran Alisyn Malek, May’s chief operations officer.
Malek, like Olson, came from the autonomous car industry. At GM Ventures, she oversaw the relationship with then-startup Cruise, which is now developing GM’s driverless car platform. Steve Vozar, May’s other cofounder and chief technical officer, was also a member of Ford’s driverless vehicle program and directed the University of Michigan’s APRIL robotics laboratory.
From the beginning, May’s purpose was to deploy a generation of autonomous driving into the genuine world, focusing on what can be imagined today instead of five or 10 years in the future. In accordance with this objective, the company has targeted consumers with desires for fixed-route transportation in residential neighborhoods, campuses and gated residential communities located geographically and smoothly.
“We’ve noticed Ford, Toyota and GM reports up close,” Olson told TechCrunch in an interview in August 2017. “Like the other big corporations in this space … everyone is looking for the shipping market in general.” That’s the $4 billion on-demand shipping problem, Uber thinks. We, by entering the genuine world, can not only build a successful business, but also enable a flow of knowledge and operational knowledge that will help us evolve and improve our systems faster than the [original device manufacturers] that are in study and progression mode.”
May chose not to design cars himself, but focused on software, low-level electromechanical formulas and sensors while relying on third-party hardware. His “secret sauce” was a set of algorithms that the company called multi-policy decision-making, whose paintings began in Olson’s lab. Instead of educating the formula by offering knowledge of many other scenarios and training it to respond to those scenarios, May’s software was designed to perceive the conditions and wait for what happens agent-to-agent.
To further differentiate, May has introduced a comprehensive fleet operation service that covers everything from day-to-day operations to maintenance and cleaning. It is also committed to competing with consumers to integrate their independent battery into their existing fleets.
“May Mobility, like Waymo, Cruise and Uber, is a complete construction of generation packages designed for the full diversity of autonomous driving scenarios, even if our first markets are more structured,” Olson told Xconomy in a March 2018 interview. Uber and Cruise have invested heavily in their equipment, however, the savings of achievable units, which do not require motor force and car protection, still seem to be in years. We plan to beat the 3 in the market with a successful product ».
May’s agile progression cycle allowed him to recruit his first clients in 2018, just a few months after its creation. But resources said it also put pressure on the company’s engineering team, which modernized a vehicle platform that had not been designed with May’s deployments in mind.
May used the transmission and chassis of the Polaris six-seater return gem, which has a maximum speed of 25 miles depending on the time, allowing it to be used on public roads without approval. But in cities like Columbus, Ohio, the new GEM has struggled to control adverse weather conditions. May had to cancel inconsistent operations on certain winter days because low weather conditions prevented GEM batteries from turning on.
Even under higher conditions, the availability of May’s fleet has a challenge as several ferries have been damaged at the same time. The driver from Providence, Rhode Island, who had a dozen cars on site, only saw a portion of the cars with passengers at one point while the rest was in maintenance.
And despite his Level Four autonomy target, May struggled to achieve self-driving in the cities where MLG was deployed. This is despite the fact that the ferries were guided through other sensors outside the vehicle fixed on traffic lighting devices that provided data such as the color of the street lighting devices.
Sources say this is the result of challenging management environments and GEM deficiencies.
After 500 hours of testing at Quonset Business Park in North Kingstown, Rhode Island, humans had to bypass GEM systems for left turns with traffic, red right turns, rain, wind, pedestrian status at pedestrian crossings, structure areas, some four-lane Array intersections with donkey backs, potholes and competitive drivers. The foliage also caused headaches because the pre-programmed direction was mapped when the trees were bare. Once the leaves and other foliage emerged, it interrupted the camera, radar and lidar sensors that helped the ferries navigate.
One source described May’s autonomy as “everywhere on the map.” While some days have noticed that autonomy rates are successful at 80% to 90%, functionality has fallen “well below” this mark.
Laurie Remias, May’s chief marketing officer, told VentureBeat that the company takes into account all the situations MLG faces when measuring the rate of autonomy. The figures are “very seasonal” and count the time of day as well as real-time traffic situations. “If it snows, [our diversity rate is] 0% because we don’t drive our cars autonomously in bad weather situations,” he said via email. “It is very important, when it comes to autonomy rates, to talk about them in context.”
Providence ferries also had to deal with one tricky factor: high-speed traffic. In some spaces where they were inconsistent with hate, drivers traveled up to 50 miles depending on the hour, far exceeding the maximum speed restriction for MONs, which did not have an air bag.
In the charging of airbags and air conditioning, the latter was only installed on GEM ferries in August 2019 due to the lack of compromised batteries, the MONs lacked a heating system. Engineers were forced to charge diesel heating to the platform, meaning they would be completely electric, one of their original promotion points.
“The industry is taking this to a point that doesn’t come to the point and is wrong,” Julia Gold, head of sustainability and innovation at the Rhode Island Department of Transportation, told EcoRI News in an interview last year. “[May] has not been able to demonstrate that it has made fully self-driving cars safer than human drivers … We pressure them to do better.”
May’s stumbles didn’t end in engineering. Inconsistent situations with demanding demands resulted in fewer riders than expected, with maximum routes, which were loose riders, who failed to reach the equilibrium point. One source estimates the losses of each deployment at more than $1 million consistent with the year.
In Columbus, where May’s GEM shuttles traveled several miles through the Scioto Mile district, the city spent an estimated $30 per passenger on one 1.2-mile loop, with May paying about $90 per passenger. As a result of this and the route’s poor autonomy rate, in August 2019, the city sent a 30-day notice letter to May threatening cancellation and legal action. Columbus is now pursuing action to recover a portion of contract money for what it alleges were service-level agreement violations.
One source described the May deployment in Columbus as more of a “centerpiece” than a functional travel route, with “tourist attractions” such as the Smart Columbus Experience Center, an exhibition corridor of the city’s efforts. While the ferries were going to have almost every day of sunlight, there was no point of participation for other people using the service, the source said.
Tellingly, Columbus opted to contract a May competitor, EasyMile, for its second autonomous route.
In Providence, where routes had to be suspended for several hours in warm climates due to the AIR conditioning problem, a coming and going problem was worried in a wing hanger with someone else coming and going. There were no injuries, however, the sensors were damaged.
And when the city and Columbus asked May to bring their cars into conformity with the Americans with Disabilities Act (ADA), the company crawled, taking weeks to deploy a second generation round trip with a wheelchair restriction formula and a retractable ramp. In Providence, he also refused the walks of the young men who did not bring an elevated or protective seat.
May announced that she had reached 100,000 races in total in August 2019, but has since shared an update.
May’s deployments were paid for with a combination of personal and public capital. For the Columbus route, contributions came from the Smart Circuit program backed by Smart Columbus, a public-personal partnership between a business organization and the city government. (In 2016, the U.S. Department of Transportation granted Columbus the designation of Smart City, offering it a $40 million infrastructure fund). Rhode Island spent at least $800,000, some of which came from the budget of the agreement the state won over the Volkswagen and Federal Highway Administration emissions scandal.
But the style was not lucrative enough for May in the long run, which led the company to look for strategic partners. When that didn’t happen, the company’s Serie A was $22 million, well below its $100 million, took drastic measures to cushion the blow.
As May’s C-Suite sought a B-series and envisaged a final extension of Serie A, they took the resolution of lowering bills to suppliers to stretch the combustion rate as much as possible. In what one source called the “perfect storm” of monetary pressure, May delayed bills to partners “as much as possible.” The most sensible mandate is to look for opportunities to stop bills, the concept is to keep invoices longer than invoices.
In a statement, Remias denied that the company did not pay suppliers, and said it “works closely together” with some of its suppliers to establish payment plans. He added that as of today, all those invoices have been “totally satisfied.”
Around the same time, May began presenting a projection for Toyota on how he could expand service from 4 cities: Providence, Detroit, Columbus, and Grand Rapids to more than 80 by 2022. The revised presentations referred to deployments in Tokyo, Toyota hometown, where May contemplates offering transportation for the now postponed Olympic Games.
“No one can guess how May would do it, ” said a source.
In an implicit vote of distrust, between December 2019 and January, May lost 3 of its 4 C-Suite members (COO, CTO and CCO), as well as those guilty of reliability and sales and marketing. The company’s product manager left last summer.
“Some workers left the company in the last year for new opportunities,” Remias said in response to questions about executive outings. “That said, as we position ourselves for the future, we have also expanded our control team and recently hired a vice president of software, vice president of other people’s operations, product manager, lead engineer, and vice president of staff.”
After the final touch of her pilot projects in Columbus and Rhode Island and the failure of planned deployments in Florida and Texas, May redoubled her efforts in Grand Rapids and Detroit with Bedrock, the genuine real estate company associated with Quicken Loans President Dan Gilbert. It was recently announced that GHSP, a company that introduces an ultraviolet C remedy that automatically disinfects vehicle air and high-contact surfaces, would get $80,000 from Michigan Economic Development’s PlanetM program to install the remedy on May’s ferries.
If May’s deployments had been easier, the company intended to explore transit routes by extending the length of its fleet to 25 according to the city. pointing to part of the U.S. 3 miles or less.
“Our ideal partners are other people who have demanding first- and last-mile situations that look for others to move from transit stops or car parks to their final destination,” Malek told Wired in December 2018. “What this provides us, essentially, is a captive audience, a well-known road network and nodes among which other people have to move.”
May’s deployment on the road after deployment is more bumpy than the maximum of others, however, some of their rivals did not do much better. In February, the National Highway Traffic Safety Administration (NHTSA) partially suspended EasyMile’s U.S. operations. In France after a passenger in Columbus was injured while traveling on one of the company’s driverless ferries. And in 2018, NHTSA suspended a separate Transdev program in Florida that sought to upgrade school buses with EasyMile vehicles.
The Department of Transportation published a prophetic report on the autonomous sector back and forth two years ago, highlighting the limited range of vehicles, the demanding procurement situations and the regulatory unpredictability that new companies have not yet addressed. “The market is small and many corporations in this sector delight little in system design and validation and vehicle production compared to classic car manufacturers,” the report’s authors wrote. “Automatic low-speed changes may not be suitable for all environments and services.”
The industry has no obvious successes, like Optimus Ride. But May’s setbacks illustrate the limited applicability of emerging technology, especially when it exceeds its capabilities.