Michigan mobility pilots: what to prove before choosing a test site
Sourced news brief · See the publication date above and original source below.
Michigan mobility founders choosing between a test track and a real-world pilot should start with the decision the next experiment must unlock. Is the team still proving that its technology behaves consistently under specified conditions, or is it ready to learn whether a customer can operate and buy it?
Mcity in Ann Arbor and Michigan Central in Detroit offer useful reference points for that choice. HVN’s analysis: match the setting to the uncertainty, then define the evidence and the next commercial decision before committing the pilot budget.
Use controlled testing to isolate a technical question
The University of Michigan describes Mcity’s test facility as an early-stage research and development environment. Its Mcity OS supports repeatable scenarios and control of infrastructure such as traffic lights and crossing signals.
For a startup, our inference is that repeatability is especially valuable when comparing software versions or diagnosing a failure. A controlled experiment can hold conditions steady while the team changes one component. It does not, by itself, establish customer demand or performance across every deployment environment.
Access also needs preparation. Mcity’s published instructions require external organizations to submit a track agreement and proof of liability insurance for approval before reserving track time. Safety orientation is required before arrival.
The startup membership page lists annual dues of $5,000, or $10,000 in-kind, for startups with revenue under $10 million. It describes priority facility use at reduced rates and automatic renewal unless canceled. These are the terms displayed on October 14, 2026; they do not establish an all-inclusive testing price. Request a scoped estimate covering the proposed test, equipment and support before budgeting.
Use a field pilot to examine the operating workflow
Michigan Central’s innovation page describes real-world testing and pilot programs developed with Newlab. It also links to a public transparency portal explaining technologies used in the community. The page is a starting point for understanding the ecosystem, not evidence that a particular startup has been accepted or that a slot is available.
One concrete example is the portal’s Derq traffic-monitoring pilot. The listing identifies a June 2023–June 2024 pilot period and marks it complete. It describes collecting information about traffic, crossing hot spots and near misses along the Michigan Avenue Corridor to inform crossing-safety recommendations.
That is a historical example, not a new 2026 opportunity. The listing explains the intended work but does not establish a measured reduction in crashes or a subsequent customer contract.
HVN’s inference: a useful field pilot needs both a technical deliverable and someone responsible for acting on it. In this kind of workflow, producing a traffic insight and getting an operator to change a decision are separate milestones.
Write a one-page pilot brief before choosing the venue
The following is HVN’s proposed planning checklist, not an application requirement from either organization:
- Decision: name the specific decision the result will support, such as another engineering iteration, a limited customer deployment or a paid expansion.
- Scope and baseline: specify the site, operating conditions, product version and current customer workflow. State what comparison would make the result meaningful.
- Evidence: agree on the metric, measurement method, minimum acceptable result and who will review it. Record failures and interventions as well as successful runs.
- Operating owner: identify who installs, monitors, maintains and stops the system, and who owns the customer-side decision after the test.
- Data: agree on collection, access, retention and permitted uses, including whether results may appear in investor or customer materials. Plan how affected users will be informed.
- Budget and exit: include integration, staff time, equipment, site costs and removal. Set a review date and an explicit proceed, revise or stop decision.
For example, a hypothetical delivery-robot pilot could measure completed routes alongside human interventions and staff time per delivery. Successful trips alone would leave the customer’s workload unanswered. The founder and operator should agree on thresholds appropriate to that deployment; this guide proposes no universal pass mark.
For an investor, ask for the original pilot brief alongside the results. A useful diligence question is whether the experiment changed a named buyer’s decision, and which uncertainty remains. Participation in a respected program and a finished demonstration do not establish paid adoption.
Put the resource in context
Our Bloom funding brief examines an earlier hardware hurdle: finding and qualifying suppliers. A pilot brief addresses the next question—what the assembled product must demonstrate for a customer.
Use the founder resource guide to identify regional support for a specific milestone, and the Michigan hub for related startup coverage.
Reporting note: this resource is based on public University of Michigan/Mcity and Michigan Central materials reviewed October 14, 2026. HVN conducted no interviews or site visits and has not independently validated pilot outcomes. The decision framework and explicitly labeled inferences are HVN’s original synthesis. Confirm current access, costs and program fit with the operator before committing resources.